Essay: Model for selection of cost estimation techniques: A review

Abstract
Over the years many costing techniques are evolved. Adaption of costing technique requires investment in time and money. Therefore, selection of costing technique is an important decision. The aim of the paper is to develop a model for selection of cost estimation technique. The model is developed from a detailed comparative analysis of different techniques. This paper also provides a detailed review of all qualitative and quantitative cost estimation techniques developed over the years. Furthermore, the results of the survey from 25 small scale industries in central India are discussed. Finally, future scope for research is identified.
Keywords: cost estimation, costing technique selection, manufacturing
1 Introduction
In today’s competitive world, Cost is one of the most influential factors in the outcome of a product or service. According to Drury (1992) cost required for two purposes: first, for financial accounting to allocate the manufacturing cost incurred during a period between the cost of goods sold and inventories; second, to provide useful information for managerial decision-making requirement (Drury & Tayles 1994). Because of this sensitive and crucial role in an organization, Cost estimation and coat optimization of products is very crucial for industry. As a result, number of research effort were taken in exploring design implications, new techniques, and methods for producing accurate and consistent cost estimates not only to generate optimum design solutions but also to achieve the maximum customer satisfaction in terms of low-cost, high-quality, and in-time product delivery.
It has been observed that over 70% of the production cost of a product is determined during the conceptual design stage. However, the design phase itself accounts for only (6%) of the total development cost (Shehab & Abdalla 2001). Once the cost is locked in the design phase, it is difficult for the manufacturer to reduce it. Therefore, it is necessary to have knowledge about the cost consequences of decisions during the planning phases of the product development cycle. In these planning phases many decisions have to be made. Therefore cost estimates in the product design process should be used to choose between design alternatives in order to make cost effective decisions.
There is not a best technique that can produce the best product cost estimation results at all stages of the new product development process (Ievtushenko & Hodge 2012). Hence appropriate selection of costing technique is must. This paper gives the comprehensive discussion for skills required for cost estimation, cost estimating procedure, factors affecting product costing and requirements for cost model. Further, the detailed analysis for selection of costing technique is given. Based on this analysis model for selection of costing technique is developed. Then authors have conduced survey to identify current practices in product costing in context of at small scale furniture manufacturing industries in India.
2 Classification of Cost
Selection of cost estimation technique requires the detail understanding of the cost. As there is wide diversity in the use of terms and concepts employed in the cost computations, this diversity is the results of variety of causes. Cost has to be classified into different categories. Hence many authors have classified the cost according to element, functions, nature or behaviour, controllability, normality, relevance of decision making and control cost in their relation to the product, account period wise, behaviour wise, cost of planning and control, cost in relation to their departments and cost for analytical purpose (Murthy & Gurusamy 2009; Dutta 2003). But classification of cost given by American Institute of Chemical Engineers (AIChE) gives better understanding of cost for selecting costing method. They have given five categories of costs, with the trend of each category being increasingly more difficult to quantify (Goh et al. 2010). The five categories of costs as follows:-
a) Direct (capital investment, recurring and nonrecurring);
b) Indirect (operating and maintenance, recurring and nonrecurring);
c) Contingent (future scenarios, accidental);
d) Intangible (customer loyalty, worker morale);
e) External costs (societal costs).
3 Skills required for cost estimation
A cost estimator plays a vital role in deciding the correct utility of the cost. For this skill and competency of the cost estimator needs to be understood and mapped for accurate and reliable cost estimation. Various researchers have explored the skills required. Rodney D. Stewart (1995) and Ievtushenko & Hodge (2012) have given different skills such as Business and finance skills, engineering and technical skill, manufacturing and assembly skill, mathematical and statistical and production planning or industrial engineering skill. Smith (1995) has given additional skills like good numerate and literate education, reasonable time spent on site, ability to read and inspect drawings, ability to communicate, facility to make accurate mathematical calculations, application of logic and common senesce, patience, sense of humor, neat methodical and tidy by habit, ability to cope with vast volume of paper, working knowledge of relevant trades, appreciation of all facets of business, curiosity, ability to question basic assumptions, confidence, close relationship with those responsible for construction, knack of picking up useful information and flexibility. Further he has given an area where estimator needs to develop knowledge and understanding. The different areas are rules of measurement relevant to project, current and anticipated market condition, needs and requirements of clients and designers, interrelationship between the resources of production and components of design, contractual obligations. The above discussion indicates that, the cost estimator should be adaptive in nature and should have good business and accounting, technical and interpersonal skills.
4 Cost Estimating Process
Rodney (1995) has defined cost estimating as the process of predicting or forecasting the cost of work activity or work output depends on input from the cost analysis activity, which is the process of studying and organizing past cost and future estimates. Further he has given 12 steps for cost estimation process. But with the evolution of new approaches many authors have modified the costing process. Blocher et al. (2006) have given six steps of for cost estimation. In NASA Cost Estimating Handbook (2008) three main parts to the NASA 12 step cost estimating process is given. Mislick & Nussbaum (2015) has given 12 steps process. Based on this, the 14 step process is presented, which is divided into three parts as shown in Fig. 1.
First part is Project Definition part. The estimator clarifies the reason for the estimate, defines expectations, and begins to understand the project that will be estimated. Once estimate is defined and data is gathered, a Work Breakdown Structure (WBS) and technical description are obtained.
The second is Cost Methodology which consists of six tasks that create the approach and framework for the estimate. Developing the ground rules and assumptions are the most revisited task in this Part. Once methodologies are selected and the data is gathered, the ground rules and assumptions, methodologies, and even the cost model are refined based on the sensitivity, risk and uncertainty analysis.
Estimate is the third part of the cost estimating process which consists of five tasks. This part includes the actual conduct, presentation, and maintenance of the cost estimate.
“Insert Fig. 1 here”
5 Classification of Costing Methods
The published literature on cost estimation covers a wide variety of costing methods. A number of researchers have attempted to categorize the cost estimation techniques using certain criteria. Drury & Tayles (1994) have compared the theory and practices of product costing and broadly classified costing methods in to full cost and variable cost for decision making. Asiedu & Gu (1998) broadly classified cost estimating models used in industry as parametric models, analogous models (estimating by analogy) and detailed models. Rush & Roy (2000) has described different cost estimating methods as Traditional cost estimating, Parametric estimating, Feature Based Costing, Neural network based cost estimation and Case based reasoning. Layer et al. (2002) classified the cost into qualitative and quantitative approach. Further quantitative cost approach is classified as statistical model, analogous model and generative-analytical model. Fixson (2004) has mention the three costing techniques as parametric, analytical, analogous. Ruffo et al. (2006) has given principal quantitative approaches to cost estimation for building the mathematical model. Principal quantitative approaches were Analogy-based techniques, parametric models, Engineering approaches, artificial neural network. Niazi et al. (2006) have first time presented an extensive hierarchical classification of estimation techniques. The classification was based on grouping the techniques with similar features into various categories.
Erkoyuncu et al. (2009) have presented different cost estimation method as estimating by analogy, activity based costing, the parametric method and extrapolation. Datta & Roy (2009) have identified different cost modelling techniques: Using industry standards/catalogues, Expert Judgments, Top-Down and Bottom Up, End to end estimating, Top down Parametric Estimates, Analogy based estimates, Modelling What-if Scenarios, Joint Cost Model. For Service costing they have mention Top down costing, Bottom-up costing/Activity based costing, Mixed Approach. Trivailo et al. (2012) have listed different costing techniques as Parametric cost estimation, Engineering build-up estimation, Analogy estimation, Expert judgment estimation, Rough order of magnitude estimation. Ievtushenko & Hodge (2012) have given classification of cost estimation methods. They have classified cost estimation techniques into qualitative and quantitative. Qualitative techniques are further classified into expert judgment, whereas quantitative techniques are classified into analogical, statistical, and analytical technique. The detail hierarchical classification of costing method is as shown in Fig. 2
“Insert Fig. 2 here”
6 Factors affecting the product costing system
Selection of costing system depends on various factors. Hence knowledge of factors affecting costing systems is required. In the literature, Drury & Tayles (1994) have given five factors affecting the product costing system:-
1) Information processing cost
2) Degree of competition faced
3) Diversity of products manufactured
4) Number of products produced
5) Proportion of overheads that cannot be directly assigned to products.
Drury & Tayles (2005) have examined the extent to which potential explanatory factors influence the level of complexity of cost systems design in UK companies. Product diversity, degree of customization, size and corporate sector were the statistically significantly most affecting variables. Al-Omiri & Drury (2007) has considered 11 factors for regression analysis that influence product costing systems. Ismail & Mahmoud (2012) have examined the organizational and environmental factors that influence the cost systems design in Egyptian manufacturing firms. Based on literature following factors affecting product costing systems are needs to be considered:-
• Information processing cost: Cost involved in the processing in information.
• Cost structure: The cost structure is measured by indirect costs as a percentage of total costs.
• Product diversity
• Firm size and Industry type
• Number of products produced
• Proportion of overheads that cannot be directly assigned to products
7 Requirements for the cost model
Over the years many cost models are developed. In order to give room for good planning, effective control and visionary decision-making, a good costing model must be selected. The information should bear certain qualities for it to be seen as capable of assisting management. Layer et al. (2002) have given requirement of a generic cost calculation model Goh et al. (2010) have mentioned the best practice for a cost model. U.S. Government Accountability Office (GAO) has given criteria for validating cost estimate. The similar criterion recommended by the Society of Cost Estimating and Analysis (SCEA) consist of following points, which are useful in selection of model
• Accuracy: This includes good curve fits and minimal error bands, based on an assessment of the most likely costs.
• Comprehensiveness: Level of detail, ground rules, and assumptions must be detailed in the documentation.
• Replicability and auditability: References to source data, significance, and goodness-of-fit statistics for cost estimation relationships (CERs) clearly detailed calculations and results and rationale for method or reference chosen.
• Traceability: This means traceable to source documentation.
The U.S. Government Accountability Office (GAO) has adopted similar criteria for validating cost estimate
8 Cost Distribution (Cost Elements)
For cost calculation it, is essential to consider all cost elements and distribute them over the product. Different authors have given the different cost distribution. Woodward (1997) divided costs into the three categories: engineering and development; production and implementation; and operation. The variation of these categories with respect time is illustrates in Fig.3.
“Insert Fig.3 here”
Further he has identified the various cost elements in detail. Asiedu & Gu (1998) have distributed the different cost occurring at product, process, logistic support level over three phases of product life cycle such as acquisition phase, utilization phase, recycling phase. Fixson (2004) has grouped different cost elements at different level such as cost elements at machine level, line level, plant level and company level. Ruffo et al. (2006) have given schema for cost model. They distributed different costs like material, production overhead, production overhead and machine into direct cost and indirect cost. Roy et al. (2011) identified cost elements for automotive industry based on the analysis of six items as delivery requirements, machine specifications, raw material specifications, standard bought out part specifications, subcontract item specifications and tool specifications. The level wise distribution of cost elements given by Fixson (2004) has better visibility and understanding. It is shown in Fig.4.
“Insert Fig.4 here”
9 Costing Techniques
Over the years many costing techniques are evolved. These techniques are described in the following paragraphs.
9.1 Qualitative Techniques
Qualitative cost estimation techniques are based on a comparison analysis of a new product with the products that have been manufactured previously in order to identify the similarities in the new one Niazi et al. (2006). Cost estimation for product similar to a past design is done by identifying similarities between old and new product with the help of historical design, manufacturing data and previous experience of an estimator. Sometimes, cost estimation can be achieved by making use of the past design and manufacturing knowledge encapsulated in a system based on rules, decision trees, etc.
9.1.1 Intuitive Technique
The intuitive cost estimation techniques are based on using the past experience. A domain expert’s knowledge is systematically used to generate cost estimates for parts and assemblies (Niazi et al. 2006). The knowledge may be stored in the form of rules, decision trees, judgments, etc., at a specific location, e.g., a database to help the end user improve the decision-making process and prepare cost estimates for new products based on certain input information.
9.1.1.1 Case Based Technique
Cased-based reasoning (CBR) approach is designed to estimate the cost of a product by assuming that similar products have similar costs. Agnar & Enric (1994) have given overview of CBR process. Tseng et al. (2005) have given five general steps for CBR algorithms as shown in Fig. 5
“Insert Fig. 5 here”
Further they used Case-Based Reasoning (CBR) algorithm to construct the new BOM. While solving a new problem, CBR technique quickly generates a right BOM that fits the present situation by retrieving previous cases. Belecheanu et al. (2003) has designed a decision support for design managers and engineers during the early phases of new product development projects. Mianaei et al. (2012) has implemented CBR in drilling of oil and gas wells in design phase in Petropars Company. Wang et al. (2008) has presented a new cost estimation concept based on the case-based reasoning (CBR) for Taiwan historical buildings. In CBR model, two retrieval techniques, ‘Inductive Indexing’ and ‘Nearest Neighbor’, are then applied to retrieve relevant cases from the knowledge-based database. Haque et al. (2000) has described describes the development and application of Case Based Reasoning (CBR) to provide decision support for project managers and engineers during the early phases of New Product Development (NPD) in a Concurrent Engineering (CE) environment. The CBR algorithms can be applied under the condition that the product feature tree is built up according to a certain specification from which an effective database is formed whereas this method avoids the lengthy and complicated procedure of delaying costing process.
The case base technique is more suitable for industries, which have similar past cases. This technique provides cost estimation despite of lack of some information. However you need enough cases in the past so that new product can be matched with previous cases. Because it is dependent on past cases, it is not suitable for highly innovative industry.
9.1.1.2 Decision Support System
These systems are helpful in evaluating design alternatives. The main purpose of these systems is to assist estimators in making better judgments and decisions at different levels of the estimation process by making use of the stored knowledge of experts in the field. Decision Support System is based on storing design, manufacturing, or other constraints as a set of rules. Since many practical situations deal with uncertainty and non-availability of heuristic data, fuzzy logic techniques are used to some extent to overcome such problems.
To incorporate experts’ experience, the artificial intelligence (AI) philosophy is used to represent and utilize a domain expert’s knowledge in a way that is oriented toward problem solving and serves as a decision-aid tool.
9.1.1.2.1 Rule Base System
These systems are based on process time and cost calculation of feasible processes from a set of available ones for the manufacture of a part based on design and/or manufacturing constraints. Rules developed in the form ‘If premises Then conclusion’ helps to select a certain type of production processes to estimate process time and cost based on parts features. Based on a set of user constraints, manufacturing processes are selected that are then used to calculate the product cost. This approach is shown diagrammatically in Fig.6. Gayretli & Abdalla (1999) develop an intelligent constraint-based system that enables designers to consider at the early stages of the design process all activities associated with product’s life cycle. Masel et al. (2009) has describes a rule-based system to quickly estimate the geometry (and thus, the volume) of the forging die that will be needed for an axisymmetric part based on its geometry.
The advantage of rule base system technique is that it is capable of providing optimized solutions by adopting the most appropriate product processes. But a product could have countless parts and features. Hence this technique is time-consuming. Therefore this system is suitable for simple and less complicated products.
“Insert Fig.6 here”
9.1.1.2.2 Fuzzy logic System
This approach to cost estimation is particularly helpful in handling uncertainty. Fuzzy rules, such as those for design and production, are applied to such problems to get more reliable estimates. Shehab & Abdalla (2002a) and Shehab & Abdalla (2002b) have explained the steps in developing fuzzy model. They have presented an example for estimating machining time of drilling hole. Hole diameter, hole depth and surface finish were considered as the input variable, while machine time was considered as output variable. A decision Table 1was prepared, which is symbolic way of representing the logical interdependence between events.
“Insert Table 1 here”
The set of rules for decision Table 1is:
Hole_Rule1
If
(The hole depth is small) AND
(The hole diameter is small) AND
(The required surface finish is normal) AND
Then
(The machining time is low)
Hole_Rule2
If
(The hole depth is small) AND
(The hole diameter is medium) AND
(The required surface finish is normal) AND
Then
(The machining time is low)
Hole_Rule3
If
(The hole depth is large) AND
(The hole diameter is medium) AND
(The required surface finish is polish) AND
Then
(The machining time is high)
Set of rules from enables the estimation of the machining time Ti for a given feature, which is multiplied by the unit time cost Ri to get the machining cost Cm, i.e.
Cm = Ri Ti
Jahan-Shahi et al. (1999) have applied fuzzy sets and probability distribution approach to solve the problem of uncertainty in cost estimation of flat plate processing industry. Mittal et al. (2010) has presented using fuzzy logic to estimate effort required in software development. Dean Ting et al. (1999) proposed new model called fuzzy multi-attribute utility (FMAU) to evaluate the cost of new technologies and affordability at early product engineering design stages.
Fuzzy logic system can be used in early design phase is the advantage of this system. Further it can be used to select the most economical assembly technique for the product in order to consider this technique during the design process and provide design improvement suggestions to simplify the assembly operations. But the system can be implemented, when you have decision table for providing means for system rule and indicating the relationship between the input and output variables of fuzzy logic. Hence the cost estimation of objects with complex features using this approach is quite tedious.