An optimization-based model for maximizing the benefits of fast-track construction activities

Abstract

Fast-track construction projects have become more popular in recent years in response to growing industry demand. By allowing downstream construction activities to start with incomplete information from upstream design activities, fast-tracking (through overlapping) allows for shorter project duration at the expense of potential rework. This leaves practitioners with the challenge of determining the optimal fast-tracking strategy to meet project schedule requirements while avoiding excessive amounts of rework. This paper presents an optimization-based model that serves as a decision support tool in scheduling fast-track construction activities. The model takes into consideration information exchange between upstream and downstream activities and uses the concepts of sensitivity and evolution to maximize the net benefits of fast-tracking. The model is illustrated on an ongoing construction project, which was analysed under various overlapping scenarios. The results indicate substantial time savings depending on the speed of evolution and sensitivity. The project schedule can be shortened by up to 50 days without causing excessive amounts of rework.

Modelling the boundaries of project fast-tracking

“…The literature discussing the implications of overlapping and rework in product development and construction projects is plentiful [26]. Recently, Dehghan and Ruwnapura [27] and Dehghan et al [20] presented a thorough review of the current and previous fasttracking models for both product development and project design and execution.…”

“…In 2012 and 2013, Khoueiry et al [26,37] developed another optimisation model for fast-tracking construction activities by calculating the probability of rework and its impact.…”

“…In this case, there would be a clear violation (as opposed to a partial override) of the associated precedence relationship. This assumption is also included in most of the recent fast-track models in order to keep the network logic intact [20,26,27,37]. A tighter limit (less overlap) can be introduced, however, if there are other constraints, for example shared resources, space or material handling, between predecessors and their successors [39].…”

“…However, the proportions attributed to overlap and rework, or the probability of rework, are not clear. Some researchers have assumed a linear relationship between cost and overlap [20,27,37], and some exponential [26] or polynomial [39], but these choices have normally been based on subjective assumptions.…”

“…However, by choosing specific combinations of α and β, homogeneous (for α=1) or heterogeneous (for α≠1) previous models' Poisson information exchange processes can also be incorporated as special cases within the proposed model [26,37,48].…”