Adjustable Cost Estimation Model for COTS-Based Development

The use of commercial off-the-shelf (COTS) products to develop systems has become increasingly prominent due to schedule and cost benefits. However, COTS integration requires considerable effort and time, and an accurate estimate of an integration cost can help a project manager plan and ensure an on-time completion of the project. The COCOTS model is a widely-used method to estimate effort in COTS-based development (CBD), but the model does not address duration estimation directly or allows for a user-customized integration process, which may differ across organizations and projects. This paper proposes an adjustable model for estimating effort and duration of the integration process of CBD. The methodology integrates three existing approaches, namely COCOTS model for cost estimation, system dynamics for software process simulation, and communication overhead assumption. The proposed framework models COTS development in three phases: 1) tailoring phase 2) glue code development phase 3) system-level integration and test phase. Contrary to the traditional models, the proposed approach allows users to adapt the integration process pattern and specify productivity factors that are appropriate for their organizations. In addition to estimating the required workforce for each time period, the proposed model also finds the minimal effort needed as well as the optimal group member allocation for the COTS integration process.