Track: Operations Research

Abstract

In the present work a Linear Programming model to optimize the costs and times incurred in a software development project considering the lines of source code as a metric of the effort made by the group of available programmers was proposed, for the modeling of the problem a limit for the execution budget was assumed, a productivity rate that reflects the real percentage of a programmer's contribution to the development of the project. The acquisition of the data for the model was done through the system called SISCOST, and through the use of surveys and statistics on the salary of the programmers as well as applications such as COCOMO II and SLOCCount to estimate the total Source lines of code of a system. Finally, the optimal amounts of participation of the different types of programmers in software development were determined and the possible implications of implementing the model are analyzed.