Optimizing time–cost trade-offs in product development projects with a multi-objective evolutionary algorithm

Time–cost trade-offs arise when organizations seek the fastest product development (PD) process subject to a predefined budget, or the lowest-cost PD process within a given project deadline. Most of the engineering and project management literature has addressed this trade-off problem solely in terms of crashing—options to trade cost for time at the individual activity level—and using acyclical networks. Previously (Meier et al. in IEEE Trans Eng Manag 62(2):237–255, 2015), we presented a rich model of the iterative (cyclical) PD process that accounts for crashing, overlapping, and stochastic activity durations and iterations. In this paper, we (1) propose an optimization strategy for the model based on a multi-objective evolutionary algorithm, called ε-MOEA, which identifies the Pareto set of best time–cost trade-off solutions, and (2) demonstrate the approach using an automotive case study. We find that, in addition to crashing, activity overlapping, process architecture, and work policy provide further managerial levers for addressing the time–cost trade-off problem. In particular, managerial work policies guide process cost and duration into particular subsets of the Pareto-optimal solutions. No work policy appeared to be superior to the others in both the cost and duration dimensions; instead, a time–cost trade-off arises due to the choice of work policy. We conclude that it is essential for managers to consider all of the key factors in combination when planning and executing PD projects.