Essay: Bipedal Robot

CHAPTER 1
INTRODUCTION
1. INTRODUCTION
1.1Bipedal Robot:
The term bipedal robot refers to a robot that walks on two legs.
In recent years there has been much interest stimulated in dynamic walking in bipedal robotics and legged locomotion in general. Part of the reason for this interest is the need for robots which can operate in human oriented environments. Humans present a very elegant model of locomotion to emulate.
Bipedal robots will operate in a human environment with much greater efficiency than any other type of robot yet devised. It is hoped that eventually bipedal robots can be used to complete tasks which are too difficult or dangerous for humans. This includes applications such as working in extreme environmental conditions (such as in fire rescue operations), with toxic gases or chemicals, with explosives (such as land mines) or as an aide to humans in similar situations. Also, a useful by-product of research into bipedal robotics will be the enhancement of prosthetic devices.
The biped walking process or biped locomotion area has been studied for a long time, but it is only in the past years, thanks to the fast development of computers, that real robots started to walk on two legs. Since then the problem has been tackled from different directions.
First, there were robots that used static walking. The control criteria were to maintain the projection of the centre- of gravity (COG) on the ground, inside of the foot support area. This approach was abandoned because only slow walking speeds could be achieved, and only on at surfaces.
1.1.1 Static Walking:
The centre of gravity of the robot is always within the area bounded by the feet that are touching the ground.
The state of research into bipedal robotics has progressed to the stage where dynamic walking gaits are being studied. Human beings usually employ a dynamic gait when walking as it is faster and more efficient than static walking [1].
Static walking assumes that the robot is statically stable. This mean that, at any time, if all motion is stopped the robot will stay indefinitely in a stable position. It is necessary that the projection of the centre of gravity of the robot on the ground must be contained within the foot support area (Figure 1.1).The support area is either the foot surface in case of one supporting leg or the minimum convex area containing both foot surfaces in case both feet are on the ground.
Figure 1.1 Static walking of bipedal robot
1.1.2 Dynamic Walking:
At significant periods during the gait the centre of gravity of the robot is outside of the area bounded by the feet that are touching the ground.
Biped dynamic walking allows the centre of gravity to be outside of the support region for limited amounts of time. There is no absolute criterion that determines whether the dynamic walking is stable or not. Indeed a walker can be designed to recover from different kinds of instabilities. However, if the robot has active ankle joints and always keeps at least one foot at on the ground then the ZMP can be used as stability criteria. The ZMP is the point where the robot’s total moment at the ground is zero. As long the ZMP is inside the support region the walking is considered dynamically stable because is the only case where the foot can control the robot’s posture. It is clear that for robots that do not continuously keep at least one foot on the ground or that do not have active ankle joints (walking on stilts), the notion of support area does not exist, and therefore the ZMP criterion cannot applied.
Dynamic walking is achieved by ensuring that the robot is always rotating around a point in the support region (Figure 1.2). If the robot rotates around a point outside the support region then this means that the supporting foot will end to get on the ground or get presses against the ground. Both cases lead to instability. To draw an analogy with static walking, if all motion is stopped then the robot will tend to rotate around the ZMP.
Dynamic walking is characterized by a small period in the walking cycle where the center of gravity of the robot is not projected vertically onto the area of either foot [2]. This requires there to be a period of controlled instability in the gait cycle, which is difficult to accomplish unless the mechanical system has been designed bearing this in mind.
Figure 1.2 Dynamic walking of bipedal robot
There have already been some bipedal robots which have realized this goal to varying degrees of success. Most notably are the Honda P-2 and P-3 robots, shown in Figure 1.3. The P-2 robot alone is reputed to have a development cost of $3 million dollars. This particular biped robot is human sized with a mass of over 200 kg. The robot can climb stairs and push objects, and incorporates a virtual-reality remote manipulator control which allows a remote operator to manipulate the gripper on the robot.
However, unless similar robots can be developed for only a fraction of the cost, they cannot be easily transformed for use in their intended application
Figure 1.3: Honda P-2 and P-3 robots.
Moreover, research institutions cannot participate in developing such technologies while they are so expensive. This thesis presents a small scale low cost bipedal walking robot which can be used to investigate control and stability issues inherent in life sized robots. We examine in detail the issues surrounding the design of such a robot as well as the development of the control software.
Static balance or static walking refers to a system which stays balanced by always keeping the center of mass (COM) of the system vertically projected over the polygon of support formed by the feet [3]. While this is the case there can be no horizontal acceleration due to tipping moments caused by gravity. Therefore whenever a foot or leg is moved, the COM must not leave the area of support formed by the feet still in contact with the ground.
Figure 1.4 describes static and dynamic walking, from an overhead viewpoint. In each subfigure the direction of motion is from left to right, and the position of the center of mass is indicated by Polygons indicate the area of support formed by the feet. For subfigure (c) the trajectory of the center of mass is indicated.
Figure 1.4 Static and dynamic walking of bipedal robot
The COM may leave the support area formed by the feet for periods of time. This allows the system to experience tipping moments, which give rise to horizontal acceleration. However, such periods of time are kept brief and strictly controlled so that the system does not become unstable. Thus one may think of a dynamically balanced system as one where small amounts of controlled instability are introduced in such a way as to maintain the overall equilibrium.
Tipping moments in one direction are negated by tipping moments in the opposite direction. When we compare the two methods of balance, we see that the static method is highly restrictive and results in movement which is slow. Very rarely do animals and humans exhibit such behavior for this reason’the velocity achievable is very low and the motion is not efficient.
However, we can see that by removing the constraining nature of the rule for static balance that the mobility of the system is increased. This is due to the increased flexibility of the movement of the legs and placement of the feet. The accelerating tipping moments can be used to achieve higher speeds, move all legs at once or to utilize footholds which are far apart.
Therefore it can be seen that in order for a bipedal robot to gain efficiency and speed, it will require dynamic balance. Much of this dissertation will be concerned with analyzing the forces on the system which result from the COM being outside the base of support of the robot. In many cases the base of support is only one foot, which gives rise to substantial accelerations possible in the horizontal plane.
1.1.3 Zero Moment Point:
One prominent class of control ideas focuses on the position of the zero moment point (ZMP), the point on the ground where the reaction force and couple have no horizontal moment component. In 2D, this is the point where the net reaction is a force with no couple (the so called center of pressure, COP). ZMP controllers focus their attention on choosing ankle torques to keep the ZMP inside the foot contact polygon, thus keeping the foot ‘at on the ground. For standing still, balance of such robots is attained primarily by manipulating the robot center of mass (COM) location with the ZMP, in e’ect chasing the COM towards the centre of the support polygon. For walking, the foot placement must be such that the ZMP can be kept inside the foot-contact polygon while the robot COM is moving on or near a desired trajectory. Although these robots may use foot placement in their balance control, the underlying principle is that of balance by ankle torques. [4]
Robots that use ZMP control for walking, most famously Honda’s ASIMO series, seem to have various characteristic attributes: they walk with bent knees that allow the controllers to have authority over all the upper body degrees of freedom; they have ‘at-bottomed feet, and they consume lots of energy, perhaps because all the robot joint angles are carefully controlled at all instances of time. [5]
1.1.4 Center of Mass:
One of the best known balance criteria is the position of the projection of the center of mass (COM) on the ground plane, which must stay within the support area. This criterion was used in other engineering disciplines before it was adopted in robotics. Robots that balance using projection of COM are called static walkers to indicate, that the static balance is always preserved. Hence, they can be safely stopped at any moment. Static walkers are rather limited in their capabilities; in particular their walk is rather slow, since they must limit their acceleration. Note that this criterion is valid only when all ground contacts of a robot lie on the same horizontal plane [6].
1.2 Lagged Locomotion:
Although legged machines seem to hold people in fascination and awe at their apparent complexity, legged robots have much more potential than their wheeled counterparts. Both types of robots are designed to travel in some pre specified environment, however it can be seen that legged robots are much more mobile. This is mainly because wheels are not suited to rough terrain, despite attempts to adapt such vehicles to these environments. While this statement may seem straightforward, it is helpful to consider exactly why.
The first reason why legged robots are more mobile is that they can use isolated footholds separated by unusable terrain to optimize support and traction. In other words, where wheels require a continuous, unbroken path of support, legged robots can traverse terrain which is discrete and discontinuous. It is this flexibility in changing the configuration of the support base that makes legged locomotion so adaptable and versatile. [3]
Legs can also function as a built-in suspension system, allowing the body to travel along a different trajectory than the feet. In this way, the path of a payload may be smooth, despite the granularity and roughness of the environment. In fact, it is this decoupling and independence of the motion of the legs and the body which gives rise to superior speed and efficiency.
While wheeled robots are undoubtedly more efficient on smooth flat surfaces, many environments deviate substantially from this ideal. In fact, as humans we often create the environments which we inhabit in such a manner that they are largely unsuitable for wheeled robots. Therefore it is natural for legged robots to be more efficient and versatile in operating in human environments. However, despite the many excellent examples of legged locomotion we have all around us, we still are a long way from producing reliable and usable legged robots and machines. By building biped robots and trying to emulate these examples, more will be learnt about the difficulties of the problem and as a result the solution will become clearer. This does not deter the research effort in this area, which is producing increasingly useful models of biped locomotion.
1.3 Nomenclature:
The various terms will be used to describe the frame of reference when describing biped robots. A three dimensional Cartesian axis is used, positioned with the origin at the base of the robot (as indicated in Figure 1.5), with the forward direction of motion proceeding along the x-axis.
The horizontal plane upon which the robot walks is referred to as the ground plane, defined by the xy-axis. Two other planes are defined’the sagittal plane, which is synonymous with the xz plane, and the lateral or frontal plane, defined by the yz-plane. In addition to this terminology, each axis is defined as a rotational axis for referring to joints. The x-axis is known as the roll axis, the y-axis as the pitch axis and the z-axis the yaw axis.
Figure 1.5: Biped walking robot link and point mass model.
In determining the mass distribution while the robot is in motion, we model each link of the robot as a rigid body with a point mass attached.
A joint with motion restricted to rotation in the plane or translation along a line is referred to as having one degree of freedom (DOF). As an illustration, in Figure 1.5 the robot drawn has seven rotational DOF on the pitch axis (at the ankle, knees and hips), and one rotational DOF on the roll axis (at the base of the trunk).
1.4 Basic elements of a robot
The basic elements components of a robot system are as or follows-
Manipulator- It consists of wrist, base & arm of the robot joined in resemblance to a human arm. Within the manipulator are the mechanical parts such as joints, pairs, transmission links, internal sensors which execute the robot movements in any number of degree of freedom. The movements of the manipulator can be explained in relation to its coordinate system which can be spherical, cylindrical, Cartesian or anthropomorphic. Movement can be servo or non-servo controlled and can be trace-to-trace motion or continuous path motion depending on the controller.
Figure 1.6 Manipulator
Controller- It is the brain of a robot. It can be as simple as a set of mechanical stops and limit switches or as complex as a complete minicomputer or microcomputer. Major functions of the controller are to store, to sequence and to position the data in memory, to initiate and stop motions of manipulator as per instructions given to interact with the environment.
Figure 1.7 Controller
Machine language instructions, i.e. program to the robot to perform the desired tasks are input through the keyboard of controller. A teach pearls may also be used for online non-textual commands. The controller converts this program into suitable signals to activate the manipulator. [7]
Power supply- The power source required for the movement of manipulator may be hydraulic, pneumatic or electrical.
Hydraulic power is the most powerful and is also the most expensive. It consists of a pump of sufficient capacity and a reservoir for hydraulic fluid. It is generally preferred for spray painting applications. Pneumatic power is the least expensive, but it also provides low power. It has limited capacity. Electric power is the most accomplished source. It can be closely controlled, thus preferred for complex and precision jobs.
End Effector- This tool is provided to interact directly with the job. Depending type of work the end effector can be gripper, a spray painting torch or a welding torch, etc. gripper may be mechanical, vacuum or magnetic type. The variety of tools and grippers which can be adopted for robot use is unlimited.
Actuators- Actuators are the muscles of the manipulators. The joints of the robots are powered by what are known as actuators that produce rotary or translator movements in the links. The power delivering system can be servo or stepper motors and hydraulic and pneumatic drives. There are also other actuators that are more novel and are used in specific situations.
Typically, electric actuators are suited for high speed and low load applications, while hydraulic actuators are suited for low speed and high load applications. The pneumatic actuators are similar to hydraulic actuators, but they are not used for high payload. The main reason that pneumatic actuators are used in industry is because shop air is easily available. However, the maximum pressure is generally 100 psi.
Transmission- The transmissions are the elements between the joints and actuators of the mechanical linkage. They are generally used for following three reasons-
1. For desirable speed and torque- The output of the actuator is not directly suitable for driving the robot linkage. For this purpose, we use appropriate gearing and transmission system.
2. For desirable form of motion- The actuator output may be kinematically different from the joint motion. To handle this situation, transmission that converts the linear motion of the actuator to the rotary motion to the elbow joint, is used.
3. For transmission of power from actuator to joint located at large distance- The actuators are generally big and heavy and often it is not practical to locate the actuators at the joint. So, it is desirable to locate them at a fixed base.
Sensors- The sensors inform the robot controller about the status of the manipulator. These are the parts that recognize the robot’s position while in movement or when static. The further action or movements depend on feedback of the information collected by the sensor. It also performs the function of gathering data about its surroundings. The sensors are essential for robots to communicate with the outside world.
The sensors used in robots can be divided into two categories-
1. Non-visual sensors- they include-
(1) Position sensors (2) velocity sensors
(3) Limit switches (4) force and tactile sensors
(5) Pressure sensors (6) acceleration sensor.
2. Visual sensors- they include-
(1) Vision sensors
(2) Touch sensors
(3) Acoustic sensors.
Software and hardware- The monitors, peripherals and computer system are the hardware parts, while operating system, robot software and system software are the software parts of the robots. The software is necessary for controllers and computer system of robot for its successful operation.
To fulfill the needs of a particular industrial task of material handling, painting, welding assembly and gluing task, robots can be made out of the above mentioned components designed and selected to suit the derived specifications. [8]
1.5 Control system of robots- Control system of robot refers to means of operating the robot’s drive system and to properly regulate its motion. There are four types of motion control systems which are used for robots. Depending upon the type of sophistication desired, the robot control systems may be of following types-
1. Limited sequence system
2. Point-to-point control with playback facility
3. Continuous path control
4. Intelligent robot
1.6 Robot configurations- There are 5 types of joints which can be used in the body and arm assembly of a robot to obtain three degrees of freedom. This gives total 53=125 possible configurations for robots. However, almost all present day commercially available robots have one of the following four configurations-
1. Cartesian coordinate configuration
2. Cylindrical coordinate configuration
3. Polar or spherical coordinate configuration
4. Articulated or jointed arm configuration
1.7 Methods of robot programming- A robot program can be defined as path in space to be followed by the manipulator, combined with peripheral actions that support the work cycle. The sequence of motions is regulated by sequencing device. Control of the motions of the endpoints is accomplished by virtue of limit switches and mechanical stops. Almost all the modern industrial robots have digital computers as their controllers with compatible storage devices as their memory units.
A robot is programmed by entering the programming commands into its controller memory. Depending upon the method used to enter the program into controller memory robot programming methods are of following types-
1. Manual method
2. Walkthrough method
3. Lead through method
4. Off-line method
Although the lead through method is the most popular method of robot programming, it has certain limitations, such as-
1. The robot cannot be used in production while it is being programmed.
2. It is not compatible with modern CAD/CAM and other information processing systems.
1.8 Robot specifications- Robot specifications include the specifications of manipulator as well as the controller.
Manipulator specification includes the specification of work envelope, load carrying capacity, speed, and type of drive used, values of control resolution, accuracy and type of end effectors used.
Controller specifications include type of programming method used, programmable features, memory capacity, software and hardware features, etc.
Various terms related with the manipulator specification are explained below-
1. Work envelope
2. Load carrying capacity
3. Speed
4. Precision of movement
5. Stability.[8]
1.9 Robot programming language- A robot programming language is used to prepare robot program off-line on a computer terminal. This program is then entered into the robot’s controller memory. Most of the robot languages in use today are a combination of textual programming and teach pendant programming. The textual language is used to define the logic and sequence of program, while specific point locations are defined using teach pendant method.
The first generation robot languages use a combination of command statements and teach pendant method for developing robot program. These languages are capable of defining manipulator motions, straight line interpolation, and branching and elementary sensor commands. The second generation languages are more efficient. These languages are capable of defining complex path, to control other devices by means of sensory data and communication and data processing.
Some commonly used robot languages are listed below
1. WAVE
2. AL
3. VAL
4. SIGLA
5. HELP
6. RAIL
7. MCL
8. ROBEX
9. AML
10. VAL-II
11. SRL
12. AUTOPASS
13. VML. [7]
1.10 Functions of robot languages- With the help of commands written in language following functions can be performed-
1. Location of robot position in the working space.
2. Point-to-point robot motion.
3. Straight line interpolation by robot.
4. Motion in a desired sequence by robot.
5. Operation of end effectors of robot.
6. The robot to respond its sensor’s signals.
7. Communicate with peripherals like printers, plotters, etc.
8. Branching of robot’s control.
9. Writing and implementation of subroutines.
10. Perform complex trajectory arithmetic computations
11. Communicate with other computer based devices and use other databases.
12. Enter data into robot control unit (RCU) by means of teach pendant.
13. Display programs on the CRT of RCU, transfer programs from storage to control and back.
14. Enter new program or edit existing programs.[8]
1.11 Applications of robot in industry- Robots in an industry are used for the following applications-
1.11.1 Parts handling- This involves tasks like-
(a) Recognizing, sorting/separating the parts.
(b) Picking and placing the parts at desired location.
(c) Loading and unloading the parts on machines.
(d) Palletizing and depalletizing.
1.11.2 Parts processing- This involves operations like-
(a) Routing
(b) Drilling
(c) Riveting
(d) Arc welding
(e) Flame cutting
(f) Debarring
(g) Grinding
(h) Spray painting
(i) Coating
(j) Sandblasting
(k) Dip coating
(l) Polishing
(m) Gluing
(n) Heat treatment.
1.11.3 Product building- Robots may be used for assembly of products like
(a) Electric motor
(b) Car bodies
(c) Solenoids
(d) Circuit boards
By operations like
1. Bolting
2. Riveting
3. Nailing
4. Fitting
5. Spot welding
6. Seam welding
7. Adhesive bonding
1.11.4 Inspection- With the increasing demand of quality control, robots can be used for 100% inspection, at very fast rates with more reliability. Robots equipped with mechanical probes, optical sensors and other measuring devices can perform dimensional checking and other forms of inspection operations. [8]
1.12 Non-industrial applications of robots- Non-Industrial applications of robots are-
1. Ocean exploration
2. Agriculture and forestry
3. Construction industry
4. Mining and cool mining
5. Defense
6. Entertainment
7. Household jobs
8. Nuclear applications
9. Medical applications
10. Space applications
CHAPTER 2
LITERATURE REVIEW
2.1 Review of Past works on the Bipedal Robot
The bipedal robot to be effectively created and use dynamic balance was developed by kato in 1983. In this robot mainly used static walking, it was termed quasi-dynamic due to a small period in the gait where the body was tipped forward to enable the robot to development forward acceleration and thus achieve a forward velocity. This achievement has largely been named as the de’ning moment where the focus of research moved from static to dynamic walking. [9]
Since in this time, advancement has been somewhat slow. The similar research group produced the WL-10RD robot which walked once more with quasi-dynamic balance in 1985 [10]. The robot was requisite to return again to static balance after the dynamic transfer of support to the opposite foot. However miura and shimoyama uncontrolled static balance entirely in 1984 when their stilt bipedal BIPER-3, which was modelled after a human walking on stilts, presented true active balance. In this concept, it limited only three actuators; one to change the angle separating the legs in the direction of motion, and the remaining two which lifted the legs out to the side in the lateral plane. Then the legs could not change length, the side actuators were used to swing the leg through without friction the foot on the walking surface. An inverted pendulum was used to plan for foot location by accounting for the accelerating tipping moments which would be produced. This three degree-of-freedom biped robot was advanced extended to the seven degree-of-freedom BIPER-4 robot. [11]
Additional approach had been taken by raibert, who established a planar hopping robot. This robot used a pneumatically driven leg for the bounding motion and was involved to a cable which restricted the motion to three degrees of freedom (pitch motion, and vertical and horizontal transformation) along a radiated path notable by the cable. A formal machine was used to track the current progress of the bounding cycle, produced by the sensor feedback. The formal machine was used to adjust the control algorithm used to ensure the constancy of the machine. A comparatively simple control system was used which modi’ed three parameters of the bounding gait, namely forward speed, foot location and body outlook. The achievement of this study encouraged raibert to range the robot and control system to hopping in three dimensions, original the area of flying ‘ight in legged locomotion. [12]
On-going through the years, a dynamic successively robot was developed by hodgins, koech-ling and raibert, covering the previous studies of one-legged hopping machines in two and three dimensions. This robot was forced to two dimensions (motion in the sagittal plane), and used a similar controller method as for the hopping robot in two-dimensions. This control system decoupled the three important control factors of body height, foot placement and body attitude, controlling these three aspects of the running gait through the use of a formal machine. The formal machine switched states when assured key feedback events occurred, and the robot was controlled differently dependent upon the current state of the system. [12, 13]
Much early exploration around this time concentrated on forcefully analytical techniques for designing and regulatory robot motion. This had the tendency to produce composite equations principal the motion of the robot, which often had no solution and had to be linearised. Sometimes this methodology was effective despite such shortcomings. kajita et al. used this method to control bipedal dynamic walking by limiting the movement of the center of mass (COM) in an perfect sense to the horizontal plane only. This motion was termed a ‘potential energy conserving orbit’ and could be expressed by a simple linear differential equation, which simpli’ed the calculations involved. [14]
Two french laboratories, LMS and INRIA Rh??ne-Alpes, have planned and built an anthropomorphic biped robot. The objectives and primary results of the plan are reported in and the execution of the postural motions and static walks achieved until now are described in [15, 16] .The current study is focused toward the generation of anthropomorphic trajectories, and toward capable ways for the biped robot to control them. The robot is ‘tted with feet prepared with sensors measuring the ground-foot forces, in order to activity the concepts of centre of pressure (COP) and zero moment point (ZMP). The idea of ZMP has been known about for extra than thirty years, but is in no way old-fashioned, then as long as gravity forces govern walking gaits, the ZMP will be a signi’cant dynamic equilibrium condition. In addition, this quite useful idea has not been completely explored, and unfortunately some misconstructions are sometimes encountered in the works. [17]
They planned a kinematic model in which the feet were in connection with the ground as shown in Figure 2.1. In order to continue the support foot stationary and let the other foot swing forward i.e. make the system shuffle, an essential condition was mentioned where the ground reaction force must operate only on the supporting foot. Their unique proposal was highly substantial due to the subsequent two facts. Mainly, it allows designing motion paths, which fulfil the mentioned condition while ignoring the atmosphere’s counter action, as the total exterior force is dynamically equilibrated with the internal forces generated by the system’s actuators. Secondly, motion design can be obtained by moment compensation in the planning because the horizontal component of the moment around the ZMP is zero (thus its name). In fact, in this initial model, the lower limbs’ motion trajectory were analytic functions, hence, the upper limb’s motion compensating the moment around the support foot was computed using an iterative analytic solution.[18]
Advanced, Vukobratovic et al. (1970) planned feedback control for the same model. As declared previously, oversight of the environment in the motion preparation can simplify the problem, but it makes the system fewer forceful against disturbances caused by slight model deviations. Consequently, the moment error should be detached using online feedback, so they planned a real-time motion modification using a sensitivity matrix whose elements measured moment versus acceleration variation, to smooth modification magnitude calculation as an opposite to the moment error.
Figure 2.1: Vukobratovic et al.’s biped shuffle kinematic model.
The repeatability and probability of the Passive Toy’s gait by engaging it on a moving treadmill. The treadmill had a fairly compliant walking surface and was prepared with a control unit, shown in Figure 2.2, for any regulations of speed and angle of incline. [19].
Figure 2.2: Completed Passive Toy.
To discover kneed passive walking as an opportunity for a robot to test machine learning. While the Passive Toy used curved feet and pitch swinging, knees over a more anthropomorphically-accurate solution for providing ground clearance and controlling step frequency. A second inspiration for discovering kneed passive walking is McGeer’s indication that bipeds with knees can, in some cases, walk more stable than those with straight legs [20].
Figure 2.3: McGeer’s kneed walker.
Now instruction to discover passive walking with knees, it activated by fabricating a robot approaching the walker built by McGeer, shown in Figure 2.3. McGeer’s robot is a two-dimensional passive walker with two couples of two legs, with one pair situated centrally and the other pair situated laterally. For both couples, the two legs are strictly joined and move synchronously, as though the robot is bipedal. The purpose of this unusual design is to horizontally balance the robot in the frontal plane. By limiting the design of this “four-legged biped” to two dimensions, the yaw effects experienced by the Passive Toy are eliminated.
A biped robot was designed and built at the Institute of System and Robotics of the Department of Electrical and Computer Engineering of the University of Coimbra, in Portugal. The mechanical structure of the robot has shown in fig. 2.4 the main joints of hip, knee, and ankle, for each leg. There is another joint, an active inverted pendulum that is used for the lateral balance of the structure. The robot carries its battery pack on this inverted pendulum. The robot is actuated by seven servo motors and the structure is made of acrylic and aluminium. It weighs 2.3 kg and is 0.5 m tall.
The robot was designed to move in both horizontal and inclined planes, to go up and down stairs, and has a speed of approximately 0.05 m/s. A 9600 bit/s RS232 wireless transmission link binds the control software that is running on a PC, to the robot. The robot board has two PIC microprocessors, one to acquire the digital values of the force sensors and the other to actuate the servo motors. [21]
Fig. 2.4. Implemented robot.
In order to create a machine capable of operating in such terrains, many advances must be made, including environment sensing, planning and guidance, e’cient and powerful actuation, high density energy storage, and gait control. The last of these, gait control, is the focus of this thesis, speci’cally for bipedal robots. Before developing control algorithms, two areas of a bipedal robotic system are considered: leg design and control architecture. Noting the success of biologic systems in navigating di’cult terrains, inspirations are taken from nature. The selected leg design has articulated hips, knees, and ankles, substantial (non-point) feet, pneumatic primary actuators, and mechanical coupling between the knee and ankle joints. The selected control architecture is a hierarchy, with general heuristics forming the top level, which are distilled into system speci’c specialized control laws and eventually into joint level actuator e’orts. A dynamic simulation method appropriate for systems with varying topology, such as legged robots, is also developed.
The study of legged robots dates to the 1960’s, when Frank and McGhee created the Phony Pony shown in Fig. 2.5, an autonomous quadruped capable of ‘creeping’ forward very slowly. Since that time, there has been an explosion of legged robots, and with each new entry into the ‘eld come new capabilities and insights. [22]
Figure 2.5: Autonomous legged robot
A constructive control design for stabilization of non-periodic trajectories of under actuated robots. An important example of such a system is an under actuated ‘dynamic walking’ biped robot traversing rough or uneven terrain. The stabilization problem is inherently challenging due to the nonlinearity, open-loop instability, hybrid (impact) dynamics, and target motions which are not known in advance. The proposed technique is to compute a transverse linearization about the desired motion: a linear impulsive system which locally represents ‘transversal’ dynamics about a target trajectory. This system is then exponentially stabilized using a modi’ed receding-horizon control design, providing exponential orbital stability of the target trajectory of the original nonlinear system. The proposed method is experimentally veri’ed using a compass-gait walker: a two-degree-of-freedom biped with hip actuation but pointed stilt-like feet. The technique is, however, very general and can be applied to a wide variety of hybrid nonlinear systems. [23]
The main components are:
1. Terrain Perception: fusion of sensors such as vision, radar, and laser, perhaps combined with pre-de’ned maps, generating a model of the terrain ahead.
2. Motion Planning: uses the terrain map, current robot state, and a model of the robot’s dynamics to plan a ‘nite-horizon feasible sequence of footstep locations and joint trajectories. Slow time-scale: motion plan might be updated once per footstep.
3. Motion Control: feedback control to stabilize the planned motion in the face of inaccurate modelling, disturbances, time delays, etc. Fast time-scale: typically of the order of milliseconds.
4. Robot State Sensing and Estimation: sensors and state estimation algorithms to provide information about the physical state of the robot to all other modules.
Figure. 2.6 Control of a walking robot.
This section contains de’nitions of some basic terms that are used in the thesis. Walking bipedal robots are robotic systems that can walk using two legs. Terms biped, which stands for any two-footed animal or robot; humanoid, which denotes mechanical systems or creatures having human appearance; and walking bipedal robot are used interchangeably in this thesis. Walking is de’ned as a locomotion of a system having multiple contacts with the ground by means of breaking and regaining these contacts without simultaneous breaking of all contacts. A typical walking cycle of a biped is shown in Figure 2.7.
The convex hull of the ground contacts in the ground plane is called the support area. In practical applications it is often represented by a polygon, hence an equivalent term Polygon of Support (POS) is often used. [24]
Figure 2.7: Walking cycle of a biped.
The phase of the walking cycle when only one foot contact with the ground is preserved is referred to as Single Support (SS), while the phase when both feet are in contact with the ground is called Double Support (DS). SS and DS are also used to denote the respective support area. Step is de’ned as a half of walking cycle, which includes one single support and adjacent double support. Footstep denotes the position of a foot in ground plane. The term ‘gait’ is used to denote a pattern of movements of a robot during walk.
In accordance with balance is de’ned as the state, in which humanoid preserves the upright position. Some authors use the word ‘stability’ instead of ‘balance’, but such use is avoided here to prevent confusion. Terms static balance and dynamic balance are used to discriminate situations when a robot is balanced while it is still, and while it is moving. The ability to preserve balance is a crucial characteristic for walking robots.
After studying many different walking robots, the Passive Dynamic Walker was selected as a model for the robot to be used in machine learning experiments. [25]
The passive dynamic walker is a type of bipedal robot that was introduced by Tad McGeer. [26] It is called dynamic” because its movement is characterized by a dynamic stability. The robot is not stable at any one point in its motion. However, it is balanced in time so that the gait is steady and smooth. All bipedal walkers, including humans, must maintain dynamic stability in order to walk without falling over. The passive dynamic walker achieves this dynamic stability through a steady rocking cycle. Passive” refers to the robot’s ability to generate locomotive movement without motor input. Instead, the passive dynamic walker can produce a steady gait using only gravity and inertia. The robot will walk stably when placed on a shallow downhill walking surface. [27]
The mechanical design of a bipedal walking robot named M2V2, as well as control strategies to be implemented for walking and balance recovery. M2V2 has 12 actuated degrees of freedom in the lower body: three at each hip, one at each knee, and two at each ankle. Each degree of freedom is powered by a force controllable Series Elastic Actuator. These actuators provide high force fidelity and low impedance, allowing for control techniques that exploit the natural dynamics of the robot. The walking and balance recovery controllers will use the concepts of Capture Points and the Capture Region in order to decide where to step. A Capture Point is a point on the ground in which a biped can step to in order to stop, and the Capture Region is the locus of such points. [28]
Yobotics is developing a twelve degree-of-freedom bipedal walking robot platform, which has been dubbed M2V2 (see Figure 2.8). The robot will be capable of high fidelity force control at each of the 12 degrees of freedom. With the ability to go beyond joint tracking trajectories, M2V2 will be used to implement, validate, and extend various bipedal control algorithms including those developed on Spring Flamingo, a planar biped, and on a simulated three dimensional robot.
.
Figure 2.8: M2V2 is a 12 DOF walking robot.
M2V2 has twelve degrees of freedom consisting of three at each hip, one at each knee, and two at each ankle, a fairly typical arrangement for bipedal walking robots. The robot does not have any upper-body degrees of freedom as its main role is in bipedal walking research. The legs are made of carbon fiber tubes permanently bonded to machined aluminium components, which define the joint ranges of motion. The body of the robot is made of seven carbon fiber plates mounted orthogonally to one another. These plates provide a stiff structure for actuator mounting as well as a protective housing for the computer, motor amplifiers and batteries. [28]
2.2 Technically Controlled Bipeds -Example
Since four decades, research institutes throughout the world have been developing bipedal robots. Despite their anthropomorphic appearance, most of the e’orts follow a more industrial approach in the design and control of their machines and apply the a fore mentioned ZMP calculation for generating joint trajectories. The most prominent representatives of this kind of robots are described in the following section, two of them in more detail.
The H7 Robot by the JSK Laboratory the Jouhou System Kougaku (jsk) Laboratory of the University of Tokyo has a long tradition of building humanoid robots, some of which are shown in Figure 2.9. The aim of its work is to develop an experimental research platform for walking, autonomous behaviour and human interaction. The design of their latest robot H7 focused on additional degrees of freedom (resulting in 30), extra joint torques, high computing power, real-time support, power autonomy, dynamic walking trajectory generation, full body motions, and three-dimensional vision support. Being 1.5 m tall and weighting 57 kg, the robot features 7 degrees of freedom per leg including an active toe joint. A real-time capable on-board computer, four lead-acid batteries, wireless lan, two ieee1394 high resolution cameras, 6-axis forces sensors and an inertial measurement unit complete the robot’s equipment [Ku’ner 01, Chestnutt 03, Nishiwaki 06].
The online walking control system of H7 allows to generate walking trajectories satisfy in a given robot translation and rotation as well as an arbitrary upper body posture. It is composed of several hierarchical layers as shown in Figure 2.9. Each layer represents a di’erent control cycle and passes its processed results to the next, lower layer which usually runs at a higher frequency.
The gait decision layer chooses the gait and calculates the footstep locations. The algorithm proposed by the authors determines the next swing leg’s foot point relative to the foot of the supporting leg. [29]
Figure 2.9: Hierarchical layers of the dynamic walking control system of the robot H7.
CHAPTER 3
PROBLEM STATEMENT
&
RESEARCH OBJECTIVES
3.1 Problem statement
The walking performance of robots can be evaluated on three important aspects:
– Robustness, i.e., the ability to handle large unexpected disturbances,
– Versatility, i.e., the ability to perform a range of different gaits,
– Energy-efficiency, i.e., the ability to consume little energy.
After about 40 years of research on walking robots, there is no bipedal robot that performs well on all of these aspects. The human outperforms its robotic counterpart by far. Figure 3.1 illustrates the performance of current bipedal robots relative to the human. One can see that a robot performs typically well on only one of three aspects. There even seems to exist a clear trade-off. For example, passive limit cycle walkers are energy-efficient but are neither versatile nor robust.
Figure 3.1: A schematic representation of the performance of various bipedal robots
The biped walking robot as a service robot in the future is supposed to work in a human living environment. When the robot moves in a complicated environment, it is highly likely for the robot to endure a variety of unknown disturbances from its surroundings. Under unexpected external force, the walking state will be changed in a short period of time, including its velocity, position, angular momentum, and etc., and the robot may diverse from its stable walking state and fall down. This work hopes to analyze the extent to which the robot can tolerate perturbations and if the perturbation exceeds the tolerance, what strategies should be adopted to overcome it and recover to stability. We hope that based on this proposed stability control system, the real biped walking robot can overcome perturbation in its walking process.
3.2 Objective of the work
The aim of this research is to design, build and implement a robot to achieve bipedal static walking and dynamic walking. While is a difficult task, especially given the severe time constraints, this has been achieved in some form. To clarify the goal, we hope to develop the robot and control system to the stage where it can walk unsupported on a flat surface indefinitely (given sufficient power supply conditions). It is hoped that the control system will be able to accept user variation of gait parameters such as step length, step period and foot lift height as a coarse basis for providing a method of controlling the speed and position of the robot.
Under ideal conditions, a model that includes gait adjustments for slight disturbances in the environment will be achieved. A real application will call for a robot to function under non-ideal conditions, and it will always be the case that the environment will vary in different locations.
Therefore the robot will be required to adjust to these fluctuations in order to maintain sufficient conditions for balance.
The research can be divided into two distinct and logically separable areas ‘ design and Implementation. The design phase refers to the research involved in creating the physical robot upon which experiments will be performed. This focuses on the devices that are used in the robot as well as physical aspects such as balance, mass distribution and forces. In contrast, the implementation phase refers to the research involved in developing a control system and preparing the robot for experimentation. This involves developing gaits, determining key control parameters and developing tools to help implement the control system and analyze collected data. In this work we will address both of these research phases in detail.
1. To design and fabricate a biped robot, with six degrees of freedom and links consisting rotary joint so as to perform desired task with greater performance.
2. To reduce human efforts by using six servo motors.
3. To improve accuracy and cycle time by biped robot capable of completing skillful tasks.
CHAPTER 4
METHODOLOGY
Bipedal robot design should be based on a design methodology that produces an appropriate mechanical structure to get the desired walk. We use a design methodology that groups passive and active walk relying on dynamic models for bipedal gait. The methodology is an iterative process, as shown in Fig.4.1. The knowledge of biped robot dynamics allow us to develop simple and efficient control systems, based on the system dynamics and not on assumptions of a simplified model, which provides valid results in simulation, but validation is difficult because of the challenge represented by the measure of position, speed or acceleration of the centre of mass in a real robotic mechanism.
4.1 PROPOSED METHODOLOGY:
1. Design a gait sequence in joint space.
2. Provide sensors and actuators at joints.
4.2 DESIGNING GAITS:
1. Controlling Balance: when standing, ‘not required’ when walking.
2. Controlling Speed: It is change step size (swing leg must keep up).
3. Controlling Height: It is used to control speed and energy efficiency.
Generate intermediate joint angles based on these constraints.
First of all we prepared a prototype of bipedal robot by using the aluminum material. The prototype is help to designing gaits for making actual model of bipedal robot. We obtain different angles of each steps of walking like human walking pattern with the help of bipedal prototype.
The following gaits to be designed with the help of prototype of bipedal robot.
Figure 4.1 first step of bipedal prototype
Figure 4.2 second step of bipedal prototype Figure 4.3 third step of bipedal prototype
Figure 4.4 forth step of bipedal prototype Figure 4.5 fifth step of bipedal prototype
4.3 METHODOLOGY FLOW CHART
NO
YES
Figure 4.6 Design methodology for biped robot design
CHAPTER 5
DESIGN
&
MODELING
5.1 DESIGN
The first task in our research project was to design the robot which would be used to investigate theories of bipedal walking. However, since the funding for the project was limited, a very cost effective design needed to be developed in order to succeed. This design was also required to withstand the rigorous of mechanical stress imposed upon it during experimentation. In this Chapter the design process will be described, important aspects of the mechanical design of the biped robot will be discussed and the final constructed robot will be presented.
5.2 MECHANICAL DESIGN
The design process involves the creation of a specification for the building of a robot upon which the chosen model of dynamic walking will be implemented. The aim is to derive the specifications such that the chosen walking model will succeed. This is not a trivial task’there are many considerations to take account of in order to ensure that the biped robot will be stable while walking. The most important of these are balance, forces, moments, torque, proportions, mass and strength.
The Mechanical design forms the basis for developing this type of walking robots. The mechanical design is divided into four phases:
1) – Determining the Mechanical constraints.
2) – Conceptual Design
3) – Building the Prototype model
4) – Specification and Fabrication of the model.
1) Determining the Mechanical Constraints
There are various design considerations when designing a Bipedal robot. Among them, the major factors that have to be considered are:-
‘ Robot Size Selection:
Robot size plays a major role. Based on this the Cost of the Project, Materials required for fabrication and the no of Actuators required can be determined. In this project miniature size of the robot is preferred so a height of 246mm is decided which includes mounting of the control circuits, but the actual size of the robot is 170mm without controlling circuits.
‘ Degrees of Freedom (D.O.F):
Human leg has got Six Degrees of freedom (Hip ‘ 3 D.O.F, Knee ‘ 1 D.O.F, Ankle ‘ 2 D.O.F), but implementing all the Six D.O.F is difficult due to increase in cost of the project and controlling of the actuators which become complex, so in this project reduced degrees of freedom is aimed so 3 D.O.F per leg has been finalized (Hip ‘ 1 D.O.F, Knee ‘ 1 D.O.F, Ankle ‘ 1 D.O.F).
‘ Link Design:
In this project we design the link by using a low cost and lightweight aluminum material called aluminum strip which is joining the servos to the leg parts wherever needed. Aluminum strip is used of various lengths according to the various lengths of the different parts of the legand it also join the servo motors to the different joints.
‘ Stability:
With Biped mechanism, only two points will be in contact with the ground surface. In order to achieve effective balance, actuator will be made to rotate in sequence and the robot structure will try to balance. If the balancing is not proper, in order to maintain the Centre of Mass, dead weight would be placed in inverted pendulum configuration with 1 D.O.F. This dead weight will be shifted from one side to the other according to the balance requirement. But in this project no such configuration is used.
‘ Foot Pad Design:
The stability of the robot is determined by the foot pad. Generally there is a concept that oversized and heavy foot pad will have more stability due to more contact area. But there is a disadvantage in using the oversized and heavy foot pad, because more material will be required leading to increased costs and no significant contribution to the stability of the system. This will also force the servo motors to apply more torque for lifting the various leg parts. By considering this disadvantage an optimal sized foot pad was used. Dimensions of the foot pad are 90X80mm.
2) Conceptual Design
Initially the Bipedal robot was conceived with ten degrees of freedom. Due to constraints faced in controlling greater number degrees of freedom we, a new design was arrived with the knowledge gathered from developing previous Bipedal models.
3) Proto type
Firstly we make a prototype modal which is made by only aluminum strip after that we made a single leg and done programming to walk it.
The new design has got Six degrees of freedom with three degrees of freedom per leg. Optimal distance was maintained between the legs to ensure that legs don’t hit each other while walking.
4) Specification and Fabrication of the model
Degrees of Freedom ‘ 3 D.O.F/Leg so total of 6 D.O.F (Hip, Knee and Ankle)
Table 5.1: Specification of Bipedal Robot
S.No. Name of Component Length (mm) Width (mm) Height (mm)
1 Length of Bipedal model 170 24 275
2 Leg length 40 43 207
3 Foot pad 80 90 3
4 Servo joint upper 40 24 40
5 Servo Clamps lower 40 3 23
Table 5.2: Weight calculation of Bipedal Robot Components
S.No. Name of Component Weight
1 Estimated servo clamp weight 60gms
2 Servo motor weight 55gms
3 Total estimated weight for a link (servomotor + servomotor bracket) 120gms
4 For 6 links (i.e. 2Legs) 720gms approx.
5 Foot pad weight (2 legs) 60gms.
6 Circuits & Batteries 350gms approx.
7 Total weight of the robot 1.180Kg approx.
5.3 DRAWING OF BIPEDAL MODEL
The 3D models are developed using AutoCAD.
Figure5.1 Orthographic view of Bipedal Robot
Figure5.2 Drawing of Bipedal Robot by using Auto CAD
Figure5.3 Static view of Bipedal Robot
Figure 5.4 Walking of Bipedal Robot
Figure 5.5 Walking of Bipedal with different angle
Figure 5.6 Final Model of Bipedal Robot
CHAPTER 6
MICROCONTROLLER ARCHITECTURE & OTHER HARDWARE
Microcontroller Architecture & Other Hardware
As mentioned in the introduction, the biped robot was designed with the aim that the controller and power supply would be on board, making the robot self-sufficient and unreliant on outside support. In this chapter we will review the electronic inventory of the robot and other hardware which is used in bipedal. First we will look at the microcontroller architecture and mechanical hardware then discuss the software development process.
Major component used in our biped modal
‘ Arduino board and controller
‘ Servo motor
‘ Aluminum strip & aluminum sheet
‘ Keypad
‘ Rely mate connector & 3mm nut bolt as per required
6.1 Arduino
Arduino is a family of single-board microcontrollers, intended to make it easier to build interactive objects or environments. The hardware consists of an open-source hardware board designed around an 8-bit Atmel AVR microcontroller or a 32-bit Atmel ARM. Current models feature a USB interface together with six analog input pins and 14 digital I/O pins that can accommodate various extension boards.
The first Arduino was introduced in 2005. Its designers sought to provide an inexpensive and easy way for hobbyists, students, and professionals to create devices that interact with their environment using sensors and actuators. Common examples for beginner hobbyists include simple robots, thermostats and motion detectors. Arduino boards come with a simple integrated development environment (IDE) that runs on regular personal computers and allows users to write programs for Arduino using C or C++.
Figure 6.1 Arduino board
Hardware
An arduino board consists of an Atmel 8-bit AVR microcontroller with complementary components that facilitate programming and incorporation into other circuits. An important aspect of the Arduino is its standard connectors, which lets users connect the CPU board to a variety of interchangeable add-on modules known as shields. Some shields communicate with the Arduino board directly over various pins, but many shields are individually addressable via an I??C serial bus, so many shields can be stacked and used in parallel. Official Arduino have used the mega AVR series of chips, specifically the ATmega8, ATmega168, ATmega328, ATmega1280, and ATmega2560. A handful of other processors have been used by Arduino compatibles. Most boards include a 5 volt linear regulator and a 16 MHz crystal oscillator (or ceramic resonator in some variants), although some designs such as the Lily Pad run at 8 MHz and dispense with the onboard voltage regulator due to specific form-factor restrictions. An arduino microcontroller is also pre-programmed with a boot loader that simplifies uploading of programs to the on-chip flash memory, compared with other devices that typically need an external programmer. This makes using an Arduino more straightforward by allowing the use of an ordinary computer as the programmer.
Software for Arduino
The Arduino integrated development environment (IDE) is a cross-platform application written in Java, and derives from the IDE for the Processing programming language and the Wiring projects. It is designed to introduce programming to artists and other newcomers unfamiliar with software development. It includes a code editor with features such as syntax highlighting, brace matching, and automatic indentation, and is also capable of compiling and uploading programs to the board with a single click. A program or code written for Arduino is called a sketch.
Arduino programs are written in C or C++. The Arduino IDE comes with a software library called “Wiring” from the original Wiring project, which makes many common input/output operations much easier. Users only need define two functions to make able to run cyclic executive program:
‘ setup(): a function run once at the start of a program that can initialize settings
‘ loop(): a function called repeatedly until the board powers off
6.2 Servo motor
A servomotor is a rotary actuator that allows for precise control of angular position, velocity and acceleration. It consists of a suitable motor coupled to a sensor for position feedback. It also requires a relatively sophisticated controller, often a dedicated module designed specifically for use with servomotors.
Servomotors are not a specific class of motor although the term servomotor is often used to refer to a motor suitable for use in a closed-loop control system.
Servomotors are used in applications such as robotics, CNC machinery or automated manufacturing.
Figure 6.2 servo motor
To fully understand how the servo works, you need to take a look under the hood. Inside there is a pretty simple set-up: a small DC motor, potentiometer an motor is attached by gears to the control wheel. As the motor rotates, the potentiometer’s resistance changes, so the control circuit can precisely regulate how much movement there is and in which direction.
When the shaft of the motor is at the desired position, power supplied to the motor is stopped. If not, the motor is turned in the appropriate direction. The desired position is sent via electrical pulses through the signal wire. The motor’s speed is proportional to the difference between its actual position and desired position. So if the motor is near the desired position, it will turn slowly, otherwise it will turn fast. This is called proportional control. This means the motor will only run as hard as necessary to accomplish the task at hand, a very efficient little guy.
Types of Servo Motors
There are two types of servo motors: AC and DC. AC servos can handle higher current surges and tend to be used in industrial machinery. DC servos are not designed for high current surges and are usually better suited for smaller applications. Generally speaking, DC motors are less expensive than their AC counterparts. These are also servo motors that have been built specifically for continuous rotation, making it an easy way to get your robot moving. They feature two ball bearings on the output shaft for reduced friction and easy access to the rest point by the adjustment of potentiometer.
Controlling of servo motors
Servos are controlled by sending an electrical pulse of variable width, or pulse width modulation (PWM), through the control wire. There is a minimum pulse, a maximum pulse and a repetition rate. A servo motor can usually only turn 90?? in either direction for a total of 180?? movement. The motor’s neutral position is defined as the position where the servo has the same amount of potential rotation in the both the clockwise or counter-clockwise direction. The PWM sent to the motor determines position of the shaft, and based on the duration of the pulse sent via the control wire the rotor will turn to the desired position. The servo motor expects to see a pulse every 20 milliseconds (ms) and the length of the pulse will determine how far the motor turns. For example, a 1.5ms pulse will make the motor turn to the 90?? position. Shorter than 1.5ms moves it to 0?? and any longer than 1.5ms will turn the servo to 180??, as diagramed below.
Figure 6.3: Pulse of Servo Motor
When these servos are commanded to move, they will move to the position and hold that position. If an external force pushes against the servo while the servo is holding a position, the servo will resist from moving out of that position. The maximum amount of force the servo can exert is called the torque rating of the servo. Servos will not hold their position forever though; the position pulse must be repeated to instruct the servo to stay in position.
Servo Motor Applications
Servos are used in radio-controlled airplanes to position control surfaces like elevators, rudders, walking a robot or operating grippers. Servo motors are small, have built-in control circuitry and have good power for the size of servomotor.
In food services and pharmaceuticals, the tools are designed to be used in harsher environments where the potential for corrosion is high due to being washed at high pressures and temperatures repeatedly to maintain strict hygiene standards. Servos are also used in in-line manufacturing, where high repetition and precise work is necessary.
6.3 Aluminum strip & aluminum sheet
Aluminum alloy 6063 is a medium strength alloy commonly referred to as an architectural alloy. It is normally used in intricate extrusions. It has a good surface finish; high corrosion resistance is suited to welding and can be easily anodized. Most commonly available as T6 temper, in the T4 condition it has good formability.
AA 6063 is an aluminum alloy, with magnesium and silicon as the alloying elements. The standard controlling its composition is maintained by The Aluminum Association. It has generally good mechanical properties and is heat treatable and weld able.
Chemical Composition
Table 6.1: Chemical composition of Aluminum Alloy 6063[30]
Elements % Present
Manganese (Mn) 0.0 – 0.10
Iron (Fe) 0.0 – 0.35
Magnesium (Mg) 0.45 – 0.90
Silicon (Si) 0.20 – 0.60
Zinc (Zn) 0.0 – 0.10
Titanium (Ti) 0.0 – 0.10
Chromium (Cr) 0.0 – 0.10
Copper (Cu) 0.0 – 0.10
Aluminum (Al) Balance
Figure 6.4 aluminum strip
Fabrication:
1. Solderability:- Good
2. Weldability:- Gas: Excellent
3. Weldability:- Arc Excellent
4. Weldability:- Resistance Excellent
5. Brazability:- Excellent
6. Workability:- Cold Average
7. Machinability:- Average
Application:
This material is used for architectural applications, shop ‘ttings, irrigation tubing, balustrading, window frames, extrusions and doors.
6.4 Keypad
We used here own designed keypad for controlling the robot. By this we move the robot also forward and backward. Our robot is also take turn right and left both side.
6.5 Rely mate connector & nut bolts
Rely mate connector is used to connect the servos to the board for power and instruction.
Nut bolt is used to join the joint and motor to each other.
Figure 6.5 connector
CHAPTER 7
GAITDEVELOPMENT
&GAIT GENERATOR SOFTWARE
7.1. GAIT DEVELOPMENT
Before a bipedal robot can walk, a gait or walking pattern must be developed for the robot to follow. There are many different ways of doing this, however the aim of gait development is to produce a gait which is dynamically stable. If the gait is dynamically stable, when the robot walks according to the gait in the absence of external disturbances, it will achieve dynamic walking. However, if the gait is not dynamically stable then the robot will fall over, since the system will be unstable.
7.1.1. WALKING GAIT
Generally walking cycle consists of two steps namely Initialization and Walking
1) Initialization:
In the Initialization step the robot will be in balanced condition and in this step the servomotors are made to return to home position. This will certainly help the robot to advance into the next step.
2) Walking:
Walking step is further classified into six phases.
Phase 1 ‘ Double Support:
In this phase both the legs are in same line and the center of mass is maintained between the two legs.
Phase 2 ‘ Single Support (Pre-Swing):
In this phase both the ankle joints are in actuated in roll orientation which shifts the center of mass towards the left leg and the right leg will be lifted up from the ground.
Phase 3 ‘ Single Support (Swing):
In this phase, the right leg is lifted further and made to swing in the air. Hip and knee joints are actuated in pitch orientation so that right leg is moved forward.
Phase 4 ‘ Post Swing:
In this phase the lifted leg is placed down with the actuation of ankle joints.
Phase 5 and 6 are the mirror image of Phase 2 and Phase 3.
After Phase 6, motion continues with a transition to Phase 1 and the walking continues.
Figure 7.1: The gait cycle'(a) shows some of the phases from the gait cycle, and these are correspondingly drawn in (b).
It takes approximately 30 seconds to complete one walking cycle (all 6 phases). Bipedal robot has a step length of approximately 10mm. The Robot has the capability of carrying a dead weight of approximately 150gms.
7.2. GAIT GENERATOR SOFTWARE
In order to study bipedal walking in robotics more effectively and efficiently, a program was developed to input, animate and simulate gaits for biped robots. Entitled the ‘Gait Generator’, it not only enabled the gait being studied to be input and processed into a form which can be tested on the biped robots, but more importantly allowed the user to visualize the gait through real-time animation. In this chapter we describe the ‘Gait Generator’ program, the under lying data requirements and capabilities of this software, and an evaluation of the usefulness and practicability of the software.
MATLAB (Matrix laboratory) is a multi-paradigm numerical computing environment and fourth-generation programming language. Developed by Math Works, MATLAB allows matrix manipulations, plotting of functions and data, implementation of algorithms, creation of user interfaces, and interfacing with programs written in other languages, including C, C++, Java, Fortran and Python.
7.2.1. Program description
This program operates on window system. The program is used to enter a walking gait for the bipedal robot, which is created by entering the angles of the robot links at different times during the gait.
The user of the program has available a number of visual controls at their disposal to aid in entering the gait. Each gait is entered frame by frame. Frames are considered as a still-shot of the angles of the robot links at a particular instant in time during the gait. At the users request, the program will animate the sequence of frames in real-time, smoothly interpolating for the time periods in between frames. This allows the user to simulate the walking gait and assess the likelihood of both the gait working correctly, and the robot walking with that particular gait.
7.2.2. Program requirements
As mentioned previously, in order for the program to run it requires a windows operating environment running the windowing system. The data requirements of the program are simply either a gait input by the user, or a previously saved gait.
The output of the system is a gait generator format file, which saves the angles of the links at each point in time, for later reloading, editing and animation. Also output are the C source files which can be compiled with other source files. These source files are compiled into assembly language by a C compiler and then cross-assembled for the At Mega-8 (AVR) microcontroller which exists in the hardware of the biped robot. This compiled byte code is downloaded to the robot via a serial link cable to the windows machine.
7.2.3. Program evaluation
This program has been of enormous benefit to the research group in developing the first simple gaits for the biped robots. The eventual aim is to prepare several simple walking gaits which can be run on the robots. When robots execute these gaits, various sensors will be used to collect data while the robot is walking. This data can be analyzed and correlated against the movements of the robot in an attempt to determine how this sensory feedback can be applied to control the robot in the future.
It is likely that as a control system is developed for the robot, this program will be used less frequently. However, in the meantime, until this stage eventuates, the Gait Generator software is an important tool in the study of bipedal walking. The tool has helped with entering, simulating and visualizing walking gaits, and has saved much time in the preparation for data collection.
7.3 PROGRAMMING OF BIPEDAL ROBOT
We prepared the program for bipedal robot by using C++ language.
// bipadal robot programme
#include
Servo myservo1,myservo2,myservo3,myservo4,myservo5,myservo6;
// create servo object to control a servo
// a maximum of eight servo objects can be created
int k1 = 8;
int k2 = 9;
int k3 = 10;
void setup()
{
pinMode(k1,INPUT);
pinMode(k2,INPUT);
pinMode(k3,INPUT);
myservo1.attach(2);
myservo2.attach(3);
myservo3.attach(4);
myservo4.attach(5);
myservo5.attach(6);
myservo6.attach(7);
myservo1.write(35);
myservo2.write(100);
myservo3.write(85);
myservo4.write(80);
myservo5.write(100);
myservo6.write(140);
digitalWrite(k1,HIGH);
digitalWrite(k2,HIGH);
digitalWrite(k3,HIGH);
delay(2000);
}
voidop_servo(int ,int );
void loop()
{
if(digitalRead(k1)==LOW)
{
walk(4);
}
else if(digitalRead(k2)==LOW)
{
left_turn();
}
else if(digitalRead(k3)==LOW)
{
right_turn();
}
// left_turn();
// while(1);
}
void walk(int steps)
{
inti;
//————-step 1——————//
myservo6.write(110);
delay(500);
myservo1.write(70);
delay(100);
myservo6.write(140);
delay(1000);
for(i=1;i<steps;i++)
{
//————-step 2——————//
myservo2.write(120);
myservo3.write(65);
myservo5.write(120);
myservo4.write(50);
myservo1.write(35);
delay(1000);
//————–step3 ————
myservo1.write(70);
delay(1000);
myservo6.write(110);
delay(1000);
myservo3.write(85);
myservo2.write(100);
myservo1.write(35);
myservo4.write(80);
delay(100);
myservo5.write(100);
delay(1000);
//——————step 4—————-//
myservo5.write(80);
myservo4.write(100);
myservo2.write(80);
myservo3.write(105);
myservo6.write(150);
delay(1000);
myservo5.write(100);
delay(500);
myservo6.write(110);
delay(500);
myservo1.write(70);
delay(100);
myservo6.write(140);
myservo2.write(100);
myservo4.write(80);
myservo3.write(85);
delay(1000);
}
myservo1.write(35);
}
voidleft_turn()
{
inti = 0;
for(i=0; i<7;i++)
{
//————-step 1——————//
myservo6.write(110);
delay(500);
myservo1.write(70);
delay(100);
myservo6.write(140);
delay(1000);
//————-step 2——————//
myservo2.write(120);
myservo3.write(65);
myservo5.write(120);
myservo4.write(50);
delay(500);
myservo1.write(35);
delay(1000);
//————–step2 ————
myservo1.write(35);
myservo2.write(100);
myservo3.write(85);
myservo4.write(80);
myservo5.write(100);
myservo6.write(140);
delay(2000);
}
}
voidright_turn()
{
inti = 0;
for(i=0; i<7;i++)
{
//————-step 1——————//
myservo1.write(70);
delay(1000);
myservo6.write(110);
delay(1000);
myservo1.write(35);
delay(1000);
//————-step 2——————//
myservo5.write(80);
myservo4.write(100);
myservo2.write(80);
myservo3.write(105);
delay(500);
myservo6.write(140);
myservo5.write(100);
delay(1000);
//————–step2 ————
myservo1.write(35);
myservo2.write(100);
myservo3.write(85);
myservo4.write(80);
myservo5.write(100);
myservo6.write(140);
delay(2000);
}
}
CHAPTER 8
RESULT
RESULT
The most important part of research and experimentation is obtaining and analyzing results to verify previous explanations and theories. During the course of this research, several experiments were completed and the results were collected. In which we move it with load and without load condition.
8.1. WALKING EXPERIMENT
8.1.1 Stationary walking on the ground plane
Due to the instability of both the intuitive and periodic function approaches, it was decided to examine in more detail the ability of the robot to balance in the lateral and sagittal plane while walking on the spot. The reason behind this that the overall stability of the walking gait depends upon the timing of the motions in both of these planes. If the timing is incorrect in either plane, it will affect the stability of the robot in the other plane as well as the overall stability of the walking gait. It is therefore desirable to study the relationship between the stability of the robot and the motions in these planes.
The stability of the robot in the lateral plane can be examined by enabling the robot to walk continuously in the same position on the ground plane. This has the desired effect of reducing tipping moments which are created when the robot translates a foot, which can increase the instability of the robot. Thus the problem of control is simplified, since the parameters of step height and step length are constant, and the motion is effectively restricted to the lateral plane. Through variation of the step period and the magnitude of the trunk motion in this plane, the timing relationship between these parameters can be examined.
8.1.2. WALKING WITH LOAD AND WITHOUT LOAD
We saw during the load position the bipedal is moving proper but it required more power and many times when load is more it was unbalanced. We put load in box which is attached beside the robot and operate it the result is below in table.
Table 8.1 A bipedal robot with loading condition.
S. No. Load / Weight (gm.) Result
01 10 Walk smoothly
02 20 Walk smoothly
03 50 Walk but slow
04 200 Walk but unbalance
05 250 Fully unbalanced
While we run it freely without load, it run smoothly and required low power.
TASKS IT CAN PERFORM
1. Walk in forward, left and right direction.
2. Walk with load and without load.
3. Walk in plane area and ground area.
TASKS IT CAN BE AIMED FOR
1. Walk and turning(left and right side) very fast.
2. To design the walking pattern for bipedal.
Major Problems Faced
1. 3-Servo Biped
‘ Length of its legs and distance between them.
‘ Surface area of feet.
‘ Length of link.
‘ Synchronization of servo.
2. 6-Servo Biped
‘ Almost all the problems faced in 3- servo biped.
‘ 3-D modal of various parts.
‘ Coding and controlling the servos.
‘ Joining the servo into the link.
‘ Setting the correct axis of the servos and deciding the correct angles to be turned by the servos.
‘ Understanding the arduino and excess of wires from servos.
‘ Balancing the Biped legs and maintaining its center of gravity.
8.2 SOLUTION
1. Many online videos and research paper is very helpful to operating the servos.
2. After a lot of trials and tests we went ahead with sticking the servos to the links and joint.
3. After a lot of trials and testing the axis and angle of each servo was individually set and then all the 6 were tested together.
4. The placement of arduino was decided as such to balance the legs and maintain the center of gravity.
CHAPTER 9
CONCLUSION & FUTURE SCOPE
9.1 CONCLUSION
It is not trivial to implement dynamic walking in bipedal robot, however in this dissertation the relevant issues for designing and constructing such a machine have been discussed, and the possibilities for implementing a control system to coordinate the dynamic gait have been examined.
A useful tool for the development of gaits for our bipedal robots has been developed. A more complex control system may be required in order to stabilize the robot sufficiently. This may require an adaptive control system, such as artificial neural networks (ANN), genetic algorithms (GA) or fuzzy logic. However, it is more likely that the method of gait generation will also need improvement, in order to generate the most stable gait before control is implemented, minimizing the control problem.
The importance of gait generation has been established, as well as the significance of control system to stabilize the robot while in motion. Both must be present for dynamic bipedal walking to succeed, and both require more research.
Research in this field is important for developing robots which can operate in normal human environments, and can adapt to disturbances and variations in the environment, enabling them to traverse over uneven terrain. In the future, with the convergence of many widely differing fields of research, this is becoming a reality.
9.2. FUTURE SCOPE
As the reader might well imagine, the scope for future work in bipedal robotics is immense and there are many different avenues of research which may be investigated.
These BIPED legs can further be aimed into creating a full humanoid. First the hands can be added enabling it to walk and pick up things, punch obstacles and other functions. Then a head can also be attached enabling it to recognize colors through image processing. More human like functions can be assigned to the robot leading it to become something bigger than only a pair of bipedal legs. A controller can also be made to manually handle the bot and make it perform tasks.
The Artificial humans and autonomous artificial servants have a long history in human culture, though the term Robot and its modern literary conception as a mobile machine equipped with an advanced artificial intelligence are more fairly in recent years. The literary role of the artificial life has evolved over time. Early myths present animated objects as instruments of divine will, later stories treat their attempted creation as a blasphemy with inevitable consequences, modern tales range from apocalyptic warnings against blind technological progress to explorations of the ethical questions raised by the possibility of sentient machines.
After some research and experimentation, the future approaches which can be undertaken and are most likely to generate increased success are more readily identified. Through attempting these new methods, it is hoped that the goal of achieving dynamic bipedal walking will be realized. This chapter will identify some areas of research which should be investigated in the future.
9.2.1. Design Improvement and its Implementation
The general design of the robot could be improved in several ways. The mass distribution of the robot is crucial as this determines the position of the center of mass (COM) of the robot, as well as the total mass. The total mass of the robot is important as this determines the maximum torque required from the servos and therefore dictates the size of the actuators which should be used. However, of more importance is the placement of the COM as this will ultimately determine the stability of the robot.
9.2.2 Gait Generation
As mentioned before, it is desirable to generate a gait that is dynamically stable. This is because such a gait will require minimal control in order to maintain the stability of the system. Several methods of gait generation were tried, with varying methods of success. One powerful method of generating a dynamically stable gait is synthesis through zero moment point (ZMP) calculation.
The ZMP is defined as the point on the ground plane where the total moment due to gravity and inertia equals zero, or upon which there is no torque acting. Another synonymous definition of the ZMP is the position where the resultant of the ground reaction force of the robot penetrates the ground plane. Therefore, by definition, the ZMP must always lie inside the support area formed by the feet in contact with the ground.
9.2.3. Motion Control
It is highly likely that a more complex control method will need to be developed and used. There are many and varied control systems which can be used, and combinations of different types of control systems can be created. One branch of powerful control teqniques involves intelligent or learning control systems, such as artificial neural networks (ANNs).
CHAPTER 10
REFERENCES
[1] H. Hemami And Y. F. Zheng, ‘Dynamics And Control Of Motion On The Ground And In The Air with Application To Biped Robots,’ Journal Of Robotic Systems, Vol. 1, No. 1, Pp. 101’116, 1984.
[2] R. E. Goddard, F. Zheng, Yuan, and H. Hemami, ‘Control Of the Heel-Off To Toe-Off Motion
Of A Dynamic Biped Gait,’ Ieee Transactions On Systems, Man, And Cybernetics, Vol. 22, No. 1, Pp. 92’102, 1992.
[3] M. H. Raibert, Legged Robots That Balance. Cambridge, Ma: Mit Press, 1986.
[4] Pranav Audhut Bhounsule, A Controller Design Framework For Bipedal Robots: Trajectory Optimization And Event-Based Stabilization, May 2012.
[5] Y. Sakagami, R. Watanabe, C. Aoyama, S. Matsunaga, N. Higaki, And K. Fujimura.The Intelligent Asimo: System Overview And Integration. In Proc. Of International Conference on Intelligent Robots and Systems, Lausanne, Switzerland, Volume 3, Pages 2478’2483, 2002.
[6] Pierre-Brice Wieber. On The Stability Of Walking Systems. In Proceedings of the International Workshop on Humanoid and Human Friendly Robotics, Tsukuba, Japon, 2002. (Cited On Page 5.)
[7] Yeralan, S., & Emery, H. (2000). Programming and interfacing the 8051 microcontroller in C and Assembly. Rigel Press.
[8] MicroElectronika. (2004). Architecture and programming of 8051 MCU (Chapter 1).
[9] T. Kato, A. Takanishi, H. Jishikawa, And I. Kato, ‘The Realization Of The Quasi-Dynamic Walking By The Biped Walking Machine,’ In Fourth SymposiumOn Theory And Practiceof Robots And Manipulators(A. Morecki, G. Bianchi, And K. Kedzior, Eds.), (Warsaw),Pp. 341’351, Polish Scienti’c Publishers, 1983.
[10] A. Takanishi, M. Ishida, Y. Yamazaki, And I. Kato, ‘The Realization Of Dynamic Walking By The Biped Walking Robot Wl-10rd,’ In Icar’85, Pp. 459’466, 1985.
[11] H. Miura And I. Shimoyama, ‘Dynamic Walk Of A Biped,’ International Journal Of Robotics Research, Vol. 3, Pp. 60’74, 1984.
[12] M. H. Raibert, Legged Robots That Balance. Cambridge, Ma: Mit Press, 1986.
[13] J. Hodgins, J. Koechling, and M. H. Raibert, ‘Running Experiments withA Planar Biped,’ In Robotics Research: The 3rd Int. Symp.(O. Faugeras and G. Giralt, Eds.), (Cambridge, Ma), Pp. 349’355, Mit Press, 1986.
[14] S. Kajita, T. Yamaura, And A. Kobayashi, ‘Dynamic Walking Control Of A Biped Robot Along A Potential Energy Conserving Orbit,’ Ieee Transactions On Robotics And Automation,pp. 431’438, Aug. 1992.
[15] P. Sardain, M. Rostami, and G. Bessonnet, ‘An Anthropomorphic Biped Robot: Dynamic Concepts and Technological Design,’ Ieee Trans. Syst.Man Cybern. A, Vol. 28, Pp. 823’838, Nov. 1998
[16] B. Espiau and P. Sardain, ‘The Antropomorphic Biped Robot Bip2000,’ In Proc. Ieee Int. Conf. Robot. Automation, San Francisco, Ca, 2000, Pp. 3997’4002
[17] C. Azevedo, ‘Control Architecture And Algorithms Of The Anthropomorphic Biped Robot Bip2000,’ In Proc. Int. Conf. Climbing Walking Robots, Madrid, Spain, 2000, Pp. 285’293
[18] Yariv Bachardeveloping Controllers for Biped Humanoid Locomotion University of Edinburgh 2004.
[19] Russ Tedrake. Robots. http://Hebb.Mit.Edu/People/Russt/.
[20] Tad Mcgeer. Passive Walking With Knees. In Ieee International Conference On Robotics And Automation, Pages 1640{1645, 1990.
[21] Jo??o P. Ferreira2, Manuel Crisostomo A. Paulo Coimbra And Bernardete Ribeiro,Svr Controller For A Biped Robot With A Humanlike Gait Subjected To External Sagittal Forces.
[22] Alexander Douglas Perkins Control Of Dynamic ManeuversFor Bipedal Robots May 2010.
[23] Ian R Manchester1, Uwe Mettin2, Fumiya Iida3and Russ Tedrake1, Stable Dynamic Walking Over Uneven Terrain.
[24] Alexander Sherikov, Model Predictive Control Of A Walking Bipedal Robot Using Online Optimization Studies From The Department Of Technology At Orebro University Orebro 2012.
[25] B. Borovac, And V. Potkonjak. Towards A Uni’ of Basic Notions and Terms in Humanoid Robotics. Robotica25(1):87’101, January 2007. (Cited On Pages 4 And 5.)
[26] Tad Mcgeer. Passive Dynamic Walking. International Journal of Robotics Research, 9(2):62{82, 1990.
[27] Ming-Fai Fong, Mechanical Design Of A Simple Bipedal Robot, Massachusetts Institute Of Technology June 2005.
[28]Jerry Pratta,Ben Kruppb, Design Of A Bipedal Walking Robot,Ainstitute Of Human And Machine Cognition, 40 South Alcaniz Street, Pensacola, FlUsa 32502 ,Byobotics, Inc., 2138 Sinton Avenue, Cincinnati, Oh Usa 45206
[29]Koichi Nishiwaki 1,*, James Kuffner 2,1, Satoshi Kagami1,3, Masayuki Inaba3 and Hirochika Inoue,The experimental humanoid robot H7:a research platform for autonomous behaviour , Phil. Trans. R. Soc. A (2007) 365, 79’107,2006.1921Published online 17 November 2006
[30] Material Data Sheet, Aluminium Alloy 6063, ThyssenKrupp Materials (UK) Ltd.
Text books:
1. SR Deb,Robotics technology & flexible automation-TMH.
2. MP Groover, Industrial Robotics- TMH.
CHAPTER 11
PUBLICATIONS
[Malviya, 4(5): May, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
(ISRA), Impact Factor: 2.114
IJESRT
INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES & RESEARCH
TECHNOLOGY
DESIGN AND DEVELOPMENT OF WALKING BIPEDAL ROBOT WITH THE HELP
OF ARDUINO CONTROLLER
Deepti Malviya*, Suman Sharma
* Truba college of Engineering & Technology, Indore, India. Truba college of Engineering & Technology, Indore, India.
ABSTRACT
A biped is a two legged walking robot which imitates human gait. It is used in places where wheels cannot go easily for example stairs, terrains, etc. It is easier for bipedal robots to exist in a human oriented environment than for other types of robots. With advances in science and technology, the interest to study the human walking has developed the demand for building the Bipedal robots. In this paper we describe the design, fabrication and analysis of Bipedal walking robot. The main objective of the project is to study about the theories and the practical challenges involved in making the bipedal ant to design the walking pattern for this. The Bipedal walking robot is designed with minimum number of actuators (servomotor) and it is controlled by the arduino based easily operated micro controller. The robot is designed by aluminum strip and simple servomotor. It walks like a human by balancing the centre of mass.
KEYWORDS: Bipedal robot, Arduino, gait and actuators.
INTRODUCTION
This paper describes the design, fabrication and analysis of Bipedal walking robot. The main objective of the project is to study about the theories and the practical challenges involved in making it. The Bipedal walking robot is designed with minimal number of actuators (RC Servomotor). Two french laboratories, LMS and INRIA Rh??ne- Alpes, have designed and constructed an anthropomorphic biped robot, Bip. The goals and initial results of the project are reported in [1] and [2] and the implementation of the postural motions and static walks achieved until now are described in [3]. When we look at terrestrial animals, we notice that most of them move around using legs – be it two, four or more legs. If we compare legged animals and non-legged ones, we often find that the legged ones are typically more agile than their non-legged counterparts – they can traverse more types of terrains. This fact has inspired many researchers to look into building legged systems as an alternative to the wheeled systems that currently dominate mobile robotics and even land transportation systems. Besides, the legged robots will be able to work in human environments better than wheeled robots can [4]. Currently, many institutions around the world are conducting research on legged systems [5, 6, 7, 8, 9, 10]. The common legged systems being researched are the two legged (biped), four legged (quadruped) and six legged (hexapod) ones. Arguably, between these, bipedal walking is the most interesting and most di??cult to achieve. Though there are many researches on bipedal walking, sadly, only a handful had succeeded in producing a stable dynamic walk. This only underscores the di??culty in developing a bipedal walking system. The development of a bipedal walking system actually consist of a variety of research areas -robotics, mechanics, electronics, control and biomechanics all contribute to various aspects of this research. System integration is therefore very crucial to the successful development of a bipedal robot. Attempts at building walking machines can be traced back at least to the 1960s. In addition to research concerning bipedal robots, efforts were also made to develop monopedal (Raibert, 1986) and quadrupedal robots (Furusho et al. 1995). One of the first functioning bipedal robots was developed in the 1970s by Kato (Kato and Tsuiki, 1972). Today, there are many bipedal robot projects in the world, and the number of active projects is growing rapidly. Here, we will briefly review some of the work in bipedal robotics to date. We will mainly focus on motor skills for walking robots. [11]
http: // http://www.ijesrt.com ?? International Journal of Engineering Sciences & Research Technology
[Malviya, 4(5): May, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
(ISRA), Impact Factor: 2.114
METHODOLOGY
In this methodology the following steps have been performed. In this synthesis of dimensions was done followed by selection of material, actuator and modeling in design software. Analysis of bipedal robot is done with the help of Arduino.
SYNTHESIS OF DIMENSION
SELECTION OF MATERIAL
SELECTION OF ACTUATOR
MODELLING IN DESIGN
SOFTWARE
ANALYSIS
SYNTHESIS OF GAINT
ELECTRONIC CIRCUIT
PREPARATION/MICROPROCESSOR
PROGRAMMING
TESTING
TROBLESHOOTING
Figure 2.1
DESIGN
The design process involves the creation of a specification for the building of a robot upon which the chosen model of dynamic walking will be implemented. The aim is to derive the specifications such that the chosen walking model will succeed. This is not a trivial task’there are many considerations to take account of in order to ensure that the biped robot will be stable while walking. The most important of these are balance, forces, moments, torque, proportions, mass and strength.
The Mechanical design forms the basis for developing this type of walking robots.
A. Determining the Mechanical constraints.
B. Conceptual Design
C. Building the Prototype model
D. Specification and Fabrication of the model
http: // http://www.ijesrt.com ?? International Journal of Engineering Sciences & Research Technology
[Malviya, 4(5): May, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
(ISRA), Impact Factor: 2.114
Determining the Mechanical Constraints
There are various design considerations when designing a Bipedal robot. Among them, the major factors that have to be considered are:-
‘ Robot Size Selection:
Robot size plays a major role. Based on this the cost of the model, materials required for fabrication and the no of actuators required can be determined. In this project miniature size of the robot is preferred so a height of 246mm is decided which includes mounting of the control circuits, but the actual size of the robot is 170mm without controlling circuits.
‘ Degrees of Freedom (D.O.F):
Human leg has got Six Degrees of freedom (Hip ‘ 3 D.O.F, Knee ‘ 1 D.O.F, Ankle ‘ 2 D.O.F), but implementing all the Six D.O.F is difficult due to increase in cost of the project and controlling of the actuators which become complex, so in this project reduced degrees of freedom is aimed so 3 D.O.F per leg has been finalized (Hip ‘ 1 D.O.F, Knee ‘ 1 D.O.F, Ankle ‘ 1 D.O.F).
‘ Link Design:
In this project we design the link by using a low cost and lightweight aluminum material called aluminum strip which is joining the servos to the leg parts wherever needed. Aluminum strip is used of various lengths according to the various lengths of the different parts of the leg and it also join the servo motors to the different joints.
‘ Stability:
With Biped mechanism, only two points will be in contact with the ground surface. In order to achieve effective balance, actuator will be made to rotate in sequence and the robot structure will try to balance. If the balancing is not proper, in order to maintain the Centre of Mass, dead weight would be placed in inverted pendulum configuration with 1 D.O.F. This dead weight will be shifted from one side to the other according to the balance requirement. But in this project no such configuration is used.
‘ Foot Pad Design:
The stability of the robot is determined by the foot pad. Generally there is a concept that oversized and heavy foot pad will have more stability due to more contact area. But there is a disadvantage in using the oversized and heavy foot pad, because more material will be required leading to increased costs and no significant contribution to the stability of the system. This will also force the servo motors to apply more torque for lifting the various leg parts. By considering this disadvantage an optimal sized foot pad was used. Dimensions of the foot pad are 90X80mm.
Conceptual Design
Initially the Bipedal robot was conceived with ten degrees of freedom. Due to constraints faced in controlling greater number degrees of freedom we, a new design was arrived with the knowledge gathered from developing previous Bipedal models. The new design has got Six degrees of freedom with three degrees of freedom per leg. Optimal distance was maintained between the legs to ensure that legs don’t hit each other while walking.
The 3D models are developed using AutoCAD.
http: // http://www.ijesrt.com ?? International Journal of Engineering Sciences & Research Technology
[Malviya, 4(5): May, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
(ISRA), Impact Factor: 2.114
Figure 3.1
Figure 3.2
Proto type
Firstly we make a prototype modal which is made by only aluminum strip after that we made a single leg and done programming to walk it.
Figure 3.3
http: // http://www.ijesrt.com ?? International Journal of Engineering Sciences & Research Technology
[Malviya, 4(5): May, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
(ISRA), Impact Factor: 2.114
Figure 3.4
Specification and Fabrication of the model
Degrees of Freedom ‘ 3 D.O.F/Leg so total of 6 D.O.F (Hip, Knee and Ankle)
Table 3.1
S.No. Name of Length Width Height
Component (mm) (mm) (mm)
1 Length of 170 24 275
Bipedal
model
2 Leg length 40 43 207
3 Foot pad 80 90 3
4 Servo joint 40 24 40
upper
5 Servo 40 3 23
Clamps
lower
Table 3.2
S.No. Name of Component Weight
1 Estimated servo clamp 60gms
weight
2 Servo motor weight 55gms
3 Total estimated weight 120gms
for a link (servomotor +
servomotor bracket)
4 For 6 links (i.e. 2Legs) 720gms
approx
5 Foot pad weight (2 legs) 60gms.
6 Circuits & Batteries 350gms
approx
http: // http://www.ijesrt.com ?? International Journal of Engineering Sciences & Research Technology
[Malviya, 4(5): May, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
(ISRA), Impact Factor: 2.114
Figure 3.5
Figure 3.6
RESULT
The most important part of research and experimentation is obtaining and analyzing results to verify previous explanations and theories. During the course of this research, several experiments were completed and the results were collected. In which we move it with load and without load condition.
WALKING WITH LOAD AND WITHOUT LOAD
We saw during the load position the bipedal is moving proper but it required more power and many times when load is more it was unbalanced. We put load in box which is attached beside the robot and operate it the result is below in table.
Table 3.3
S. Load / Weight Result
No. (gm.)
01 10 Walk smoothly
02 20 Walk smoothly
03 50 Walk but slow
04 200 Walk but unbalance
05 250 Fully unbalanced
http: // http://www.ijesrt.com
?? International Journal of Engineering Sciences & Research Technology
[Malviya, 4(5): May, 2015] ISSN: 2277-9655
(I2OR), Publication Impact Factor: 3.785
(ISRA), Impact Factor: 2.114
While we run it freely without load, it run smoothly and required low power.
CONCLUSION
It is not trivial to implement dynamic walking in bipedal robot, however in this dissertation the relevant issues for designing and constructing such a machine have been discussed, and the possibilities for implementing a control system to coordinate the dynamic gait have been examined.
A useful tool for the development of gaits for our bipedal robots has been developed. A more complex control system may be required in order to stabilize the robot sufficiently. This may require an adaptive control system, such as artificial neural networks (ANN), genetic algorithms (GA) or fuzzy logic. However, it is more likely that the method of gait generation will also need improvement, in order to generate the most stable gait before control is implemented, minimizing the control problem.
The importance of gait generation has been established, as well as the significance of control system to stabilize the robot while in motion. Both must be present for dynamic bipedal walking to succeed, and both require more research. Research in this field is important for developing robots which can operate in normal human environments, and can adapt to disturbances and variations in the environment, enabling them to traverse over uneven terrain. In the future, with the convergence of many widely differing fields of research, this is becoming a reality.
REFERENCES
[1] P. Sardain, M. Rostami, and G. Bessonnet, ‘An anthropomorphic biped robot: Dynamic concepts and technological design,’ IEEE Trans. Syst. Man Cybern. A, vol. 28, pp. 823’838, Nov. 1998.
[2] B. Espiau and P. Sardain, ‘The antropomorphic biped robot BIP2000,’ in Proc. IEEE Int. Conf. Robot.
Automation, San Francisco, CA, 2000, pp. 3997’4002.
[3] C. Azevedo, ‘Control architecture and algorithms of the anthropomorphic biped robot Bip2000,’ in Proc.
Int. Conf. ClimbingWalking Robots, Madrid, Spain, 2000, pp. 285’293.
[4] D.W. Seward, A. Bradshaw, and F. Margrave. The anatomy of a humanoid robot. Robotica, 14:437’445, 1996.
[5] P. Sardain, M. Rostami, and G. Bessonnet. An anthropomorphic biped robot: Dynamic concepts and technological design. IEEE Trans. on Systems,Man, and Cybernetics – Part A: Systems and Humans, 28(6):823’838, November 1998.
[6] H. Lim and A. Takanishi. Waseda biped humanoid robots realizing human-like motion. Proc. of the 6th Intl. Workshop on Advanced Motion Control, pages 525’530, 2000.
[7] K. Hirai, M. Hirose, Y. Haikawa, and Takenaka. T. The development of honda humanoid robot. Proc. Of the 1998 IEEE Intl. Conf. on Robotics and Automation, pages 1321’1326, May 1998.
[8] A. Konno, N. Kato, T. Shirata, S. Furuta, and M. Uchiyama. Development of a light-weight biped humanoid robot. Proc. of the 2000 IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems, pages 1565′ 1570, 2000.
[9] F. Delcomyn and M.E. Nelson. Architectures for a biomimetic hexapod robot. Robotics and Autonomou Systems, 30:5’15, 2000.
[10] T. Furuta, T. Tawara, Y. Okumura, M.Shimizu, and K. Tomiyama. Design and construction of a series of compact humanoid robots and development of biped walk control strategies. Robotics and Autonomous Systems, 37:81’100, 2000.
[11] MattiasWahde and Jimmy Pettersson. A BRIEF REVIEW OF BIPEDAL ROBOTICS RESEARCH Chalmers University of Technology,Division of Mechatronics, Dept. of Machine and Vehicle Systems,SE 412 96 G??oteborg, Sweden.
http: // http://www.ijesrt.com ?? International Journal of Engineering Sciences & Research Technology
[Gupta, 5(2): April-June, 2015] ISSN: 2277-5528
Impact Factor: 3.145
INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES
& MANAGEMENT
DESIGN & CONSTRUCTION OF WALKING BIPEDAL ROBOT
Deepti Malviya1* Suman Sharma2
Truba college of Engineering & Technology, Indore 1,2
ABSTRACT
It is easier for bipedal robots to exist in a human oriented environment than for other types of robots. Furthermore, dynamic walking is more efficient than static walking. For a biped robot achieve dynamic balance while walking, a dynamic gait must be developed. Two different approaches to gait generation are presented’an intuitive approach using software for gait animation, and a periodic approach that provides a scalable gait with parameters for controlling step length, step height and step period.
Despite several decades of research, locomotion of bipedal robots is still far from achieving the graceful motions and the dexterity observed in human walking. Most of today’s bipeds are controlled by analytical approaches based on multi body dynamics, pre-calculated joint trajectories, and Zero-Moment Point considerations to ensure stability. However, beside their considerable achievements these methods show several drawbacks like strong model dependency, high energetic and computational costs, and vulnerability to unknown disturbances. In contrast to this, human locomotion is elegant, highly robust, fast, and energy e’cient. These facts gave rise to the main hypothesis of this thesis, namely that a control system based on insights into human motion control can yield human-like walking capabilities in two-legged robots.
INTRODUCTION
Bipedal robots will operate in a human environment with much greater efficiency than anyother type of robot yet devised. It is hoped that eventually bipedal robots can be used to complete tasks which are too difficult or dangerous for humans. This includes applications such as working in extreme environmental conditions (such as in fire rescue operations), with toxic gases or chemicals, with explosives (such as land mines) or as an aide to humans in similar situations. Also, a useful by-product of research into bipedal robotics will be the enhancement of prosthetic devices.
The state of research into bipedal robotics has progressed to the stage where dynamic walking gaits are being studied. Human beings usually employ a dynamic gait when walking as it is faster and more efficient than static walking [1].
Figure 1.1 Static walking of bipedal robot
Dynamic walking is characterised by a small period in the walking cycle where the centerof gravity of the robot is not projected vertically onto the area of either foot [2]. This requiresthere to be a period of controlled instability in the gait cycle, which is difficult to accomplish unless the mechanical system has been designed bearing this in mind.
Figure 1.2 dynamic walking of bipedal robot
Attempts at building walking machines can be traced back at least to the 1960s. In additionto research concerning bipedal robots, efforts were also made to develop monopedal (Raibert, 1986) and quadrupedal robots (Furusho et al. 1995). One of the first functioning bipedal robots was developed in the 1970s by Kato (Kato and Tsuiki, 1972). Today, there are many bipedal robot projects in the world, and the number of active projects is growing rapidly. Here, we will briefly review some of the work in bipedal robotics to date. We will mainly focus on motor skills for walking robots. [3]
LITERATURE REVIEW
Int. J. of Engg. Sci. & Mgmt. (IJESM), Vol. 5, Issue 2: April-June: 2015, 79-83
[Gupta, 5(2): April-June, 2015]
There are several good reasons for developing bipedal walking robots, despite the factthat it is technically more difficult to implement algorithms for reliable locomotion in such robots than in e.g. wheeled robots. First, bipedal robots are able to move in areas that are normally inaccessible to wheeled robots, such as stairs and areas littered with obstacles that make wheeled locomotion impossible. Second, walking robots cause less damage on the ground than wheeled robots. Third, it may be easier for people to interact with walking robots with a humanoid shape rather than robots with a nonhuman shape (Brooks, 1996). It is also easier for a (full-scale) humanoid robot to function in areas designed for people (e.g. houses, factories), since its humanlike shape would allow it to reach shelves etc.
The first biped robot to be successfully created and use dynamic balance was developed by Kato in 1983 [4]. While this robot largely used static walking, it was termed quasi-dynamic due to asmall period in the gait where the body was tipped forward to enable the robot to gain forward acceleration and thus achieve a forward velocity. This achievement has largely been cited as thedefining moment where the focus of research shifted from static to dynamic walking.
Since this time, progress has been somewhat sluggish. The same research group produced the WL-10RD robot which walked once more with quasi-dynamic balance in 1985 [5]. The robot was required to return again to static balance after the dynamic transfer of support to the opposite foot. However Miura and Shimoyama [6] abandoned static balance entirely in 1984 when their stilt biped BIPER-3, which was modelled after a human walking on stilts, showed true active balance. Simple in concept, it contained only three actuators; one to change the angle separating the legs in the direction of motion, and the remaining two which lifted the legs out to the side in the lateral plane. Since the legs could not change length, the side actuators were used to swing the leg through without scuffing the foot on the walking surface. An inverted pendulum was used to plan for foot placement by accounting for the accelerating tipping moments which would be produced. This three degree-of-freedom robot was later extended to the seven degree-of-freedom BIPER-4 robot.
Another approach had been taken by Raibert [7], who developed a planar hopping robot. This robot used a pneumatically driven leg for the hopping
ISSN: 2277-5528
Impact Factor: 3.145
motion and was attached to a tetherwhich restricted the motion to three degrees of freedom (pitch motion, and vertical and horizontaltranslation) along a radial path inscribed by the tether. A state machine was used to trackthe current progress of the hopping cycle, triggered by the sensor feedback. The state machinewas then used to modify the control algorithm used to ensure the stability of the machine. Arelatively simple control system was used which modified three parameters of the hopping gait,namely forward speed, foot placement and body attitude. The success of this research motivated Raibert to extend the robot and control system to hopping in three dimensions, pioneering the area of ballistic flight in legged locomotion.
EXAMPLES FOR TECHNICALLY CONTROLLED BIPEDS
Since four decades, reseach institutes throughout the world have been developing bipedal robots. Despite their anthropomorphic appearance, most of the e’orts follow a more industrial approach in the design and control of their machines and apply the aforementioned zmp calculation for generating joint trajectories. The most prominent representatives of this kind of robots are described in the following section, two of them in more detail.
The H7 Robot by the JSK Laboratory The Jouhou System Kougaku (jsk) Laboratory of the University of Tokyo has a long tradition of building humanoid robots, some of which are shown in Figure 2.7. The aim of its work is to develop an experimental research platform for walking, autonomous behaviour and human interaction. The design of their latest robot H7 focused on additional degrees of freedom (resulting in 30), extra joint torques, high computing power, real-time support, power autonomy, dynamic walking trajectory generation, full body motions, and three-dimensional vision support. Being 1.5 m tall and weighing 57 kg, the robot features 7 degrees of freedom per leg including an active toe joint. A real-time capable on-board computer, four lead-acid batteries, wireless lan, two ieee1394 high resolution cameras, 6-axis forces sensors and an inertial measurement unit complete the robot’s equipment
[Ku’ner 01, Chestnutt 03, Nishiwaki 06].
The online walking control system of H7 allows to generate walking trajectories satisfyin a given robot translation and rotation as well as an arbitrary upper body posture. It is composed of several hierarchical layers as shown in Figure 2.6. Each layer represents a di’erent control cycle and passes its processed
Int. J. of Engg. Sci. & Mgmt. (IJESM), Vol. 5, Issue 2: April-June: 2015, 79-83
[Gupta, 5(2): April-June, 2015]
results to the next, lower layer whichusually runs at a higher frequency.
The gait decision layer chooses the gait and calculates the footstep locations. The algorithm proposed by the authors determines the next swing leg’s foot point relative to the foot of the supporting leg.
Figure 2.1: dynamic walking control system of the robot H7.
Methodology
3.1ProposedMethodology:
Design a gait sequence in joint space.
Provide sensors and actuators at joints.
3.2 Designing Gaits:
1. Controlling Balance: when standing, ‘not required’ when walking.
2. Controlling Speed: It is change step size (swing leg must keep up).
3. Controlling Height: It is used to control speed and energy efficiency.
4. Generate intermediate joint angles based on these constraints.
Mechanical Design
The design process involves the creation of a specification for the building of a robot upon which the chosen model of dynamic walking will be implemented. The aim is to derive the specifications such that the chosen walking model will succeed. This is not a trivial task’there are many considerations to take account of in order to ensure that the biped robot will be stable while walking. The most important of these are balance, forces, moments, torque, proportions, mass and strength.
ISSN: 2277-5528
Impact Factor: 3.145
The Mechanical design forms the basis for developing this type of walking robots. The mechanical design is divided into four phases:
A – Determining the Mechanical constraints.
B – Conceptual Design
C – Building the Prototype model
D – Specification and Fabrication of the model.
Firstly we make a prototype modal which is made by only aluminum strip after that we made a single leg and done programming to walk it.
Figure 4.1 A Prototype Model of Bipedal Robot
The 3D models are developed using AutoCAD.
Figure 4.1 A Model of Bipedal Robot
Figure 4.2A Model of Bipedal Robot
Int. J. of Engg. Sci. & Mgmt. (IJESM), Vol. 5, Issue 2: April-June: 2015, 79-83
[Gupta, 5(2): April-June, 2015]
Figure 4.3 Final Model of Bipedal Robot
Specification and Fabrication of the model
Degrees of Freedom ‘Total of 6 D.O.F (Hip, Knee and Ankle)
Name of Lengt Widt Heig
S.N ht
Componen h h
o. (mm
t (mm) (mm)
)
Length of
1 Bipedal 170 24 275
model
2 Leg length 40 43 207
3 Foot pad 80 90 3
4 Servo joint 40 24 40
upper
Servo
5 Clamps 40 3 23
lower
Table 4.1 Dimension of Bipedal Model
ISSN: 2277-5528
Impact Factor: 3.145
S.No. Name of Component Weight
1 Estimated servo clamp 60gms
weight
2 Servo motor weight 55gms
Total estimated weight
3 for a link (servomotor + 120gms
servomotor bracket)
4 For 6 links (i.e. 720gms
2Legeights) approx
5 Foot pad weight (2 legs) 60gms.
6 Circuits & Batteries 350gms
approx
7 Total weight of the robot 1.180Kg
approx
Table 4.2 Weight Calculation of all Components
RESULT
Stationary walking on the ground plane:
The stability of the robot in the lateral plane can be examined by enabling the robot to walk continuously in the same position on the ground plane. This has the desired effect of reducing tipping moments which are created when the robot translates a foot, which can increase the instability of the robot. Thus the problem of control is simplified, since the parameters of step height and step length are constant, and the motion is effectively restricted to the lateral plane. Through variation of the step period and the magnitude of the trunk motion in this plane, the timing relationship between these parameters can be examined.
Walking with Load and without Load:
We saw during the load position the bipedal is moving proper but it required more power and many times when load is more it was unbalanced. We put load in box which is attached beside the robot and operate it the result is below in table
Int. J. of Engg. Sci. & Mgmt. (IJESM), Vol. 5, Issue 2: April-June: 2015, 79-83
[Gupta, 5(2): April-June, 2015]
S. No. Load / Result
Weight
(gm.)
01 10 Walk smoothly
02 20 Walk smoothly
03 50 Walk but slow
04 200 Walk but
unbalance
05 250 Fully unbalanced
Table 5.1: Load/Weight Calculation
While we run it freely without load, it run smoothly and required low power.
CONCLUSION
A useful tool for the development of gaits for our bipedal robots has been developed. A more complex control system may be required in order to stabilize the robot sufficiently. This may require an adaptive control system, such as artificial neural networks (ANN), genetic
algorithms (GA) or fuzzy logic. However, it is more likely that the method of gait generation
will also need improvement, in order to generate the most stable gait before control is implemented, minimizing the control problem.
The importance of gait generation has been established, as well as the significance of control system to stabilize the robot while in motion. Both must be present for dynamic bipedal walking to succeed, and both require more research.
Research in this field is important for developing robots which can operate in normal human environments, and can adapt to disturbances and variations in the environment, enabling them to traverse over uneven terrain. In the future, with the convergence of many widely differing fields of research, this is becoming a reality.
REFERENCES
[1] H. Hemami and Y. F. Zheng, ‘Dynamics and control of motion on the ground and in the air with application to biped robots,’ Journal of Robotic Systems, vol. 1, no. 1, pp. 101’116,
1984.
[2] R. E. Goddard, F. Zheng, Yuan, and H. Hemami, ‘Control of the heel-off to toe-off motion of a dynamic biped gait,’ IEEE Transactions on Systems,
ISSN: 2277-5528
Impact Factor: 3.145
Man, and Cybernetics, vol. 22, no. 1, pp. 92’102, 1992.
[3] Mattias Wahde and Jimmy Pettersson, A BRIEF REVIEW OF BIPEDAL ROBOTICS RESEARCH, Chalmers University of Technology, Division of Mechatronics, Dept. of Machine and Vehicle Systems.
[4] T. Kato, A. Takanishi, H. Jishikawa, and I. Kato, ‘The realization of the quasi-dynamic walking by the biped walking machine,’ in Fourth Symposium on Theory and Practiceof Robots and Manipulators
(A. Morecki, G. Bianchi, and K. Kedzior, eds.), (Warsaw),pp. 341’351, Polish Scientific Publishers,
1983.
[5] A. Takanishi, M. Ishida, Y. Yamazaki, and I.
Kato, ‘The realization of dynamic walking by the biped walking robot WL-10RD,’ in ICAR’85, pp. 459’466, 1985.
[6] H. Miura and I. Shimoyama, ‘Dynamic walk of a biped,’ International Journal of Robotics Research, vol. 3, pp. 60’74, 1984.
[7] M. H. Raibert, Legged Robots That Balance. Cambridge, MA: MIT Press, 1986
Int. J. of Engg. Sci. & Mgmt. (IJESM), Vol. 5, Issue 2: April-June: 2015, 79-83