Executive Summary-
Google maps enjoy monopoly from long act as common man’s map mostly essential in routing and business listing. As a matter of fact the company is going through rough patch in handling the current operational process so as to maintain high quality map and making effective change in current operational way of how the maps are getting outdated even for latest release of maps. So a research is required for analysing the current operation of Google maps and to suggest changes which can help in overcoming these shortcomings.
The current operation that Google Map Company is facing includes- privacy and outdated data. This problem has been identified and was being accepted by Google Inc. Now this list includes- very long operational process making maps redundant which lead to rejection of the map by end user. Also incomplete maps too add to existing problem where as the company was facing privacy issue from many countries in the globe. This project main goal is to dig deep and come with some optimal solution which can be easily adopted by client.
The business case of this project revolves around generating different alternatives and solution that can help in determining how the time taken for updating Google Maps in current situation can be reduced to ensure that the privacy and security needs of the community can be better addressed.
Background Company Introduction-
Google is a desktop and mobile based web application which is based on web mapping technique. The application offers real time traffic data, street view, satellite images, roads and alternative routes planning for users travelling on cycle, foot, bike (currently in beta testing) own vehicle or public transport. The maps data are updated by Google on regular basis still the latest images available were not less than 3 years. Much of these images were captured from flight cameras (birds view) as well as from satellite. The map service is widely used from all countries throughout the world and at a given time it was estimated that nearly 62% cell phone customers uses Google maps application once in week. Google Maps is basically a desktop, versatile web mapping and mobile based administration application and engineering which is being developed by Google INC, the application offers satellite symbolism, road maps, and Street View points of view, and in addition capacities, for example, a course organizer for going by foot, auto, bike (beta test), or with open transportation. Likewise upheld are maps implanted on outsider sites through the Google Maps API, and a locator for urban organizations and different associations in various nations far and wide. Satellite pictures are not overhauled progressively; notwithstanding, Google adds information to their Primary Database all the time, and the latest of these pictures are three years old.. Google Maps for versatile is the world's most mainstream application for cell phones, with in excess of 54% of worldwide cell phone holders utilizing it at any rate once amid the month of August 2013. Here as per the latest finding related to this project one of the major challenge the Google map company is facing – the time taken for maps to get updated data and the privacy issue which arises due to tracking of all users and taking the images of sensitive areas in different countries.
Below is list of all Google maps growth on Yearly basis:
2005 – First introduced in Feb 8th 2005, in starting it only supports IE and Mozilla browser. Later it was upgraded for Opera and safari.
2006 – Google started providing services to USA mainland, Japan, UK, Canada, Puerto rico. Google also added geocoding feature which is a long-time demand from developers.
2007 – Google traffic was launched to automatically load real time traffic data from many cities in USA. The same year company added driving direction support of this application.
2008 – On this year Google allows end user to edit listings of business/home address and adding new places. In August, Google signed a deal under that GeoEye would provide satellite based images.
2009 – On Mar Street View was introduced in the UK and the Netherlands. In May same year Google Maps logo was launched.
2012 – This year the map company starts mapping streams & channels with river and canal trust.
Google almost covered 250,000 miles roads and street information in the US.
2013- Google introduces Google map engine Lite which can easily be used for easy installation. During same year Google successfully covered around fourteen countries covering 3.6 million miles distance route map.
2014- This year, Street View was propelled in Cook Islands, enhancing the scope of Google Maps' 360-degree mapping symbolism to 58 nations. Amid the same time period, Google likewise finished the "biggest street view" by adding photographs of in excess of 370,000 miles (600,000 km) of roadways.
2015- On this year Google disabled user updating for map marker. Now each and every edits are passed
through review process.
2016- Google in order to enhance its feature and service, acquired analytics start-up mapping company Urban Engine. Company also rolled out 700 trillion pixels of images and data from across world, these are basically satellite images.
Reference- Golijan, R. (2012, October 11). Google updates more than 250k miles of road in Street View maps. NBC News.
Overall objective – To determine how the time taken for currently updating Google Maps can be reduced to ensure that the privacy and security needs of the application can be better addressed.
RQ1: What is the current approach used for updating Google Maps and where are the delays occurring? (Theory used- Root cause analysis)
The major areas due to which delay is happening in Google maps includes- very long operational process which is making the maps redundant, incomplete maps which could be due to insufficient or half information about location routes, delay also results from data collection and proper organization. Since collecting a huge amount of data from different sources and then analysing the data for its correctness and completeness is a herculean task.
Current Process of updating data in Google maps-
Below two processes are the current main delays causing process. They took almost 2 years for a country like Australia for getting updated.
a. Data Collection (Geo EYE, TeleAtlas, NASA, Country Governments):
Developing and updating maps for entire work needs huge amount of data mining and these cause lots of time. Three major sources of Google Maps initial data are Geo EYE, tele-atlas and NASA respective Country governments. According to Google’s report it has about 20 PB of data, which equals to 20,000 TB.
b. Software – Organizing Data & Merge/Dividing Data – Create Map-able maps:
Not only collection of data is huddle here but also analysing and merging of data equals important since data collection from satellite imagery is not much clear due to atmospheric condition. Once relevant data for maps is collected all the maps are then merged or segregated to give out a relevant map of 50KM radius.
c. Labelling-
This process stage accounts for highest number of manpower consumption. Nearly 300 employees were outsources from different part of globe with involves cheap labour cost mainly from India and china. They are high skilled trained professional who works on raw data processing. The average speed is 40km of map data labelling is done per data.
d. Quality Checking-
One of major factor of product popularity is quality. In Google maps, quality is one of the biggest challenges the company is facing. Since Google maps are getting flooded with daily reviews, edits by end users whenever new maps are introduced which causes doubling effort and duplication of same work. Hence a large Quality checking team consisting of about 1000 to 2000 analyst quality check for end user Maps.
e. Meta Quality Check & Dispute Handling-
Before sending the map to software team they passes through another quality control process called Meta quality check which essentially deals in dispute handling situation. Also the disputes happen between Labelling and QC teams during issues related to mapping. On completion, Full Mapped Maps are sent back to software team
f. Software End- Merging Data for Pushing to Live Maps:
Here once the map reaches the software team, the maps gets pushed to live maps. Every day Google maps updates many places- since its operations spread in entire world.
Reference- Takashi, O. (1995). The 5S's: Five keys to a Total Quality Environment. US: Asian Productivity Organization: Asian.
RQ2: How can the stages which consume excessive time are improved to reduce the time taken? (Fish bone and pareto chart).
Solution for decreasing the overall time is:
a. Lean Sync of Basic mapping process aimed at reducing the time to reach end users:
As discussed earlier the entire mapping process takes about 3 years to map. Major step in elimination of problems would be reduction of mapping time. A lot of time is considerably lost before and after maps are labelled. Software team should try and standardize most maps into areas depending upon their complexity rather than KM to push maps faster during labelling. An example:
b. Dedicating entire resources on quality map making:
No competition or very little competition in the field of Mapping has let Google treble around with quality at the same time spending most of its resources on research. We urge Google to finish the work which was started first before starting the next. With multiple projects running at the same time, Google has got struck in the middle of all processes which are wasting useful resources. Soon Google has to realize it and should try and get eir basic maps correct first and then build further processes on them.
c. Create a Map Maintenance team to manage changes in real world
As mentioned in the previous diagram, Road mapping is a continuous process hence, a special team – Maintenance Team could be formed which specializes in updating and keeping track of existing the maps which reduces most of the rejections. None of the processes in mapping looks into maintenance as they rely on end users to edit maps for themselves in Mapmaker tool (Assessable to everyone). This creates a problem of authentication which has to be looked into carefully.
Maintenance team can be created out of QC Team. Since the rejection rate is low, it would be suitable to pull out 100 Experts for the Maintenance Team.
d. Standardization of Maps:
One of basic challenge is standardization of maps across the world. Currently taking example like USA has the best maps with better street view updated on 2014 second is Australia with latest street view from 2011. Now these inconsistencies of data for different region add to toughness for Google work. Standardization of entire world map is a costliest work, but implementing this would certainly provide an leading edge to Google maps.
e. Select and Plan Solution:
All above are potential solutions which could improve the quality of Google maps. But the fact that most errors are inducting due to long running of the project, so a maintenance team would fit as a best solution. Major advantages of this are:
• Old maps get regularly updated.
• Maintenance team would provide strong support for other processes as well.
• Error density could be reduced.
• Authentication process of User enterable data could be completed soon.
• Helps in overall cost reduction.
f. Using IoT based sensors at roads and traffic signals for collecting real time data
While the use of sensors and AI in managing the roads and traffic could be costly but it can be seen as a onetime investment and this usage would cut down the error level to 90 percent. Here the data about traffic density, new roads and closed/alternative roads can not only be collected but they could also be updated directly to Google maps system rather than waiting for confirmation process which tends to eat maximum amount of time. Here use of Augmented reality could bring lot of new facilities like mapping could be done faster thus reducing the time of update.
Reference- https://web.archive.org/web/20131203003513/http://www.technicamix.com/2011/10/18/how-often-is-google-maps-and-google-earth-updated/
RQ3: What barriers are likely to be faced in implementing the proposed approach and how can these be overcome? (Fish bone and root cause analysis)
Some of the roadblocks for implementing the proposed approach include-
a. Moving resources from research team to quality mapping team. This could results in decreasing the research effort of organization and if Google plans to hire more resource it could increase the company overall budget.
b. Using IoT based sensors across several countries could eat a major part of company expense and could burden in overall profit.
To overcome above barriers:
a. Company should plan to invest more on increasing the overall quality of the maps and not on initial investment since once this challenge is being solved, company could generate lost of revenue by increasing the customer base.
b. Again better management of resource needs to be done since many resources are getting wasted for unnecessary process.
Situation Analysis
Google maps are currently facing very challenging situation which could also risk its existence. The privacy beach and data security is major concern for many countries since various terrorist and unlawful organizations are easily getting access of vital locations and views. One other problem includes unavailability of updated data with Google maps. This factor also contributes in making the product outdated. Some of the internal factors which the company is facing include – very long operational procedure for updating the maps which is mainly due to multi layer verification and updating which is involved for approving any new changes.
Proposed theory etc – for each RQ what theories tools etc you plan to use?
For analysing the current process and delay process in Google maps – root cause analysis & Pareto chart analysis has been planned to use. Since root cause would actually take us to the actual root of the problem and this would help in actually designing the solution for problem statement. Whereas Pareto chart would be helpful in determining the most recurring problem or we can say frequent reason for Google maps rejection in quality control.
For RQ3- In order to get the details of barriers and roadblocks related to this initiative – Force field analysis is being used to know how many of forces are against and how many are in favour of this change.
Reference- Wang, D. (2007, February 28). "Struck in the Traffic?" Google Blogger. Retrieved from Google: http://googleblog.blogspot.ca/2007/02/stuck-in-traffic.html
Flow Chart of Google maps current process-
Current Processing for how Google maps gets updated- Currently Google partner with comprehensive and authoritative data sources via its Base Map Partner program. A large number of agencies submit vector data to Google these agencies can be various countries councils, geographic surveys, yellow pages and satellite images to name some. This data is used to demonstrate changing boundaries, waterways displaying new bike, car and other path. Second source of Google maps information includes OCR (Optical character recognition) which provides capabilities to Google for reading things like road sign, traffic signs and business names. In order to show the real time traffic data Google partners with Android phones to track the owners real time data and process it to show number of people currently present it that path.
i. Root-cause analysis- This techniques could be used effectively to uncover the underlying problem. A very long operational structure of Google maps has been the main reason behind this frequent reverts of the maps. It takes at least 3 years for a map to go online which makes it pretty old and this causes most of the rejections.
While digging more deep it was found that there is a long operational process of Google maps contributes to the reason behind this frequent reverts of the maps. It takes roughly 3 years for a map to go live which makes it pretty outdated and this causes most of the rejections from end users. In addition to it, uncompleted maps make up about 35% of the issues and concern being raised by maps user. Google maps is such a multilayer project that, to complete a map, Google has to update – traffic singles, route mapping, entity and business listings, street view, house address etc. With such big diversification of the Maps service, Google is unable to nail down its 100% resources to one activity. Moreover sometimes inaccurate data given by the third Party and daily addition of new route and structure roads adds more challenging to keep up with Maps updated.
Flow chart for current process, root cause analysis
Geo-Eye NASA Yellow pages &
Tele-Atlas Governments
(Input Raw Maps) (Permissions & Inputs)
Operations Flow before Changes
BPR, Flow Chart for the new process
Operation flow future state:
Geo-Eye&NASA Governments (Input Raw Maps) IoT sensor (Permissions & Inputs)
RQ3: Questionnaire survey among staff involved in updating Google Maps.
i. Survey and questionnaire
A detailed survey has been planned which includes collecting quantitative data that can be analysed for identifying some of the major challenges which is getting acted as a roadblock for the Google employees. Some of the sample question includes:
o Out of below stages which stage is taking more time?
1. Labelling
2. Data collection
3. Software – Organizing data.
4. Quality checking
5. Dispute handling
6. Merging and publishing data
o Out of following sources of input data, which source is likely to take more time for data retrieval?
1. Data from government agencies.
2. Third party system like TeleAtlas.
3. GEO EYE
4. NASA
5. Real time traffic data & satellite data capturing
o Out of below sources, which data source is taking more time for data serialization?
1. Data from government sources like yellow pages.
2. Data from third party.
3. Satellite data and real time data capturing.
4. Others, please specify
o In current process where Artificial Intelligence can be used best to improve performance?
1. Labelling
2. Data collection
3. Software – Organizing data.
4. Quality checking
5. Merging and publishing data
o Which is best way so that the current mapping time be reduced:
1. Data retrieval is made faster so as to sync the image loading?
2. Image loading process is delayed till the data is loaded?
3. Replacing manual effort with AI for data mapping.
4. Others, please specify
o What is the most optimum way to utilize resources of company?
1. Allocating more resources at critical stages
2. Inducting efficient resources at critical stages
3. Utilize resources in fixing the basic issues which improves performance
4. Others
o How can current rework time be minimized to increase efficiency?
1. Doing a data quality check before processing starts.
2. Eliminating business rules which are likely to cause rework.
3. Others
o How can overall cost of Google map process be reduced?
1. By automating the identified manual process
2. By using AI for data analytics and processing
3. By running all process paralleling.
4. Others
o Do you think Quality Control could be strengthened by collecting data directly from IoT devices?
1. Yes
2. No
ii. Fish bone diagram
Reference- https://www.surveymonkey.com/r/Preview/?sm=z_2FB4_2BOB055SO05VsXuPY2sfJzPXiOStpEV3C78iWUxm2Yz4gUud_2F_2BrQQjXtZICJi
iii. Pareto Chart Analysis
Pareto chart analysis gives the root cause, the most recurring problem or frequent reason for rejection in quality control. This theorem is used in this project for finding problem and then defining solution. Here one of the solutions could be using of “IoT technologies” like installing electronics gadgets for capturing traffic data and route. Rejection of Maps on a daily basis has been collected:
Problems Count Cumulative Count Cumulative %
Rejection of Maps 100 100 16.2601626
Out Dated Maps 320 410 65.04065041
Incomplete Maps 140 450 89.43089431
Mis-Leading Maps 40 550 97.56097561
Mistakes by Operators 10 610 99.18699187
Others 5 615 100
Reference- https://services.google.com/fb/forms/mapmakerdatadownload/
Implementation of the Changes suggested
For changing the current operational process the 5S’s for verification of work is being considered. This 5S’s defines best organization of work and work place for increasing efficiency and scope. Here proper implementation of process takes front seat as it enhances decision making and bringing standardization in work. Introducing Maintenance team would bring 5S’s
5S’s are – Shine, Sort, Standardize, and Sustain, Straighten.
& 2S’s are – Safety and Security.
Shine or Sweep- Google maps currently allow users to edit and add maps. Millions are updates comes daily which needs authentication. Currently reviewing all edits are done and check in’s of these maps causes errors. So removing data which needs authentication could be better idea.
Sort- Sorting out old redundant map is important to omit the frequently arising errors. Introducing maintenance team should be good force to do this. Other uses could be:
• Removal of old maps with new once data comes.
• No interference with other process reducing disturbance.
• Error accumulation prevented.
• Quality increased.
Standardize- Dedicated teams are required for maintaining high quality and ordered maps, as rest of work force are mapped for mapping remaining world. An maintenance team could bring high quality maps as output.
Sustain- Sustaining with quality of maps is what Google is currently struggling with. Sooner or later they have to realize the same and should keep updating themselves regularly to match with changing world, especially when they are moving to developing countries.
Straighten- Here maintenance team could help in streamlining the error in order of error sensitiveness, this could also reduce legal problem which might arise. Additionally,
• Maps can be sorted based on oldest to newest queue.
• This promotes first come map gets first change approach.
• Simplifying the process.
Other S factors – Security & Safety- Google maps are assessable to all without any verification or validation process, this could a headache since terrorist and antisocial elements could use the map for wrong purpose. Especially street views are in concerning since privacy issue related due to uncensored paranoiac image coming directly from Google cameras. Best solution should be blurring the face object.
Proposed Data Resources
The project will further need more data and information such as –
• Survey of some existing employees of Google.
• Company data and statistics of 2016.
• Formal and informal Interviews with Google employees.
• Blogs, seminars and social media medium.
• Internal project related information from employees.
Conclusion:
Google map is currently one of most readily used application due to its numerous advantages. Here just storing 20 PB of data is quite a big challenge, maintaining and processing of data, keeping it update for relevant countries, states is even more herculean task. The operation involved in currently is simple but takes more time due to many reasons specified above. With no much competition from near industry Google could take time for implementation of process which could increase the efficiency even if this adds some extra budget to current company expedition. Since with increase in number of doubts and errors, questions were being asked regarding creditability of maps and Google could introduce a maintenance team for getting rid of errors.
Google should try and standardize most maps into areas depending upon their complexity rather than KM to push maps faster during labelling. Also Google should finish the work which was started first before starting the next. With multiple projects running at the same time, Google has got struck in the middle of all processes which are wasting useful resources. Secondly as we know road mapping is a continuous process hence, a special team – Maintenance Team could be formed which specializes in updating and keeping track of existing the maps which reduces most of the rejections. Now these inconsistencies of data for different region add to toughness for Google work. Also Google should move towards usage of sensors and AI in managing the roads and traffic could be costly but it can be seen as a onetime investment and this usage would cut down the error level to 90 percent.
References:
• Fitzsimmons, J. F. (2010). Service Management: Operations,Startegy, Information Technology. Singapore: McGraw Hill.
• Golijan, R. (2012, October 11). Google updates more than 250k miles of road in Street View maps. NBC News.
• Hart, K. H. (1989). Quantitative Methods for Quality Improvement,. Milwaukee: ASQC Quality Press.
• McLeish, D. (2005). Google Developers. Retrieved from Google Maps: https://developers.google.com/maps/
• Sieczkowski, C. (2013, January 29). Retrieved from http://www.huffingtonpost.com/2013/01/29/google-maps-north-korea-prison-camps-nuclear-complex-gulags_n_2573238.html
• Takashi, O. (1995). The 5S's: Five keys to a Total Quality Environment. US: Asian Productivity Organization: Asian.
• Wang, D. (2007, February 28). "Struck in the Traffic?" Google Blogger. Retrieved from Google: http://googleblog.blogspot.ca/2007/02/stuck-in-traffic.html
• https://web.archive.org/web/20131203003513/http://www.technicamix.com/2011/10/18/how-often-is-google-maps-and-google-earth-updated/