Service family cost estimation based on case-based reasoning
Tay, Choon Khai
Date of Issue2017
School of Mechanical and Aerospace Engineering
In recent years, companies and organisations are competing to offer high varieties of customised services to increase revenues and customers’ satisfaction. One of the effective ways to create service variations is to adopt product family design theories and methodologies. Consequently, novel service design concepts such as service family and service platform have been developed. However, research work related to service family costing is found to be scarce. Furthermore, existing service cost estimation methodologies are not feasible for estimating the costs of a service family. This research bridges these gaps by proposing a novel service family cost estimation methodology based on activity-based costing (ABC), case-based reasoning (CBR) and product modularity. The cost estimation and analysis of a service family is carried out via a three steps cost estimation framework. In step one, a service family is defined by a set of selected services. Then, activities related to these services are identified by activity diagram. A service family is regarded as a complex service delivery system. Increasing numbers of service activities and variations will complicate the process of cost estimation and analysis. To minimise the impacts, a modularisation mechanism based on product modularity is proposed to decompose a set of service family into a set of service modules. In last step, a two-stage cost estimation mechanism which is based on ABC and CBR is used to estimate service family costs. The objective of ABC is to setup cost linkages between cost elements of a service family so that accurate cost allocations can be carried out. Despite the advantages, implementation of ABC is time consuming and expensive. Furthermore, accuracy of ABC is questionable when erroneous cost allocation bases are used. Therefore, CBR is integrated to the ABC framework to overcome these shortcomings. Based on past experiences obtained from past services, activity cost is estimated without the need of cost driver measurements and time consuming data collections. Besides, service reference cluster (SRC) concept is introduced to enhance the accuracy of cost estimation. Instead of relying on single past experience, a SRC which consists of a set of similar past experiences are used to provide more rational and unbiased cost estimations. In stage two cost estimation, estimated service activities costs are allocated to relevant services. The applicability of proposed cost estimation methodology is demonstrated with a case study to estimate the costs for a fleet of long range aircraft. The result of the case study shows that, proposed cost estimation methodology is feasible to estimate the costs for a service family. It is also proves that CBR can be applied in estimating service activity costs.