A GA-based heuristic approach for offshore structure construction spatial scheduling under uncertainty

Spatial scheduling is the key process that constraints the productivity and time in offshore structure construction. However, block processing times are usually uncertain in reality. To address the uncertain characteristics of block processing times and the effects on construction makespan, a heuristic genetic algorithm approach was proposed. This paper presents a fuzzy triangular model to represent the uncertain processing times. We not only focused on the optimisation of block layouts and construction makespan, but also investigated the effects between the fuzzy makespan and the uncertainty of it. Real data are used in our empirical study to support the model. Sensitive analysis by changing uncertainty spread was conducted and results revealed that the uncertainty of fuzzy makespan imposed a non-linear negative effect on makespan. In real application, practitioners can assign greater weight to processing uncertainty when task is time oriented or less weight when task is just-in-time oriented.