Our team has been studying the design of modular construction supply chains, with emphasis on performance under conditions of operational uncertainty. The outcomes of our research have been featured in our recent articles published in the Journal of Automation in Construction ,
The benefits of modular construction stem from the use of factories for the production of building components within controlled environments. Such projects, however, are still exposed to operational delays, which may occur as a result of extreme weather events, missed deliveries, human errors and equipment failures.
Most current construction supply chain modelling tools are based on discrete event simulation (DES) models, and adopt a what-if analysis approach, instead of holistically considering uncertainty across the entire range of potential causes of disruptions.
To mitigate this gap, we developed a series of stochastic programming and robust optimisation models that could be used to determine risk-averse and highly coordinated construction logistics strategies.
1. Analysis of schedule delays
We create a statistical model of the most prevalent assembly delay factors (i.e. weather, worker productivity, transport delays, crane failures)
2. Assembly scenario generation
Using the resulting multi-factor delay model, we generate a complete set of demand scenarios and their probability of occurrence.
3. Model execution
A stochastic/robust optimisation model is used to identify the logistics plan that minimises operational cost and delay exposure.
4. Supply chain configuration
The output of the mathematical model is used to inform production rates, inventory levels and transportation schemes.
The resulting models accommodate the presence of stochastic site demands and recommend the optimal manufacturing quantities and timings, as well as transportation plans between factories, warehouses and sites.