Sustainable vehicle route planning under uncertainty for modular integrated construction: multi-trip time-dependent VRP with time windows and data analytics

Abstract

Modular integrated construction (MiC) is an innovative construction technology that boosts automation in the construction industry by shifting most of the on-site activities to controlled production facilities. However, transporting heavy, bulky, and tailor-made MiC modules to the construction site poses economic, environmental, and social challenges. Effective transportation planning is crucial to mitigate these challenges. The vehicle routing problem (VRP) is of central importance to logistics companies in determining the optimal routes for MiC module transportation. Existing literature lacks comprehensive studies on VRP that thoroughly consider the unique aspects of MiC transportation, including the need for multi-trips of trucks between the factory and the construction site, traffic conditions, and other environmental and social impacts (e.g., carbon emissions, noise, accidents, and congestion). Neglecting these factors jeopardizes the efficiency of MiC module transportation, potentially leading to project delays and undermining the sustainability benefits of MiC. Therefore, the main objective of this study is to develop a VRP model that adequately accounts for most MiC characteristics, facilitating efficient MiC module transportation. This can be achieved by proposing a new variant for the VRP model, called a multi-trip time-dependent vehicle routing problem with time windows, uncertain unloading time, and environmental and social considerations (MT-TVRPTW-UES). The MT-TVRPTW-UES is modeled as a mixed integer linear programming model. A neural network-based algorithm is utilized to predict uncertain unloading times. Additionally, we develop an ant colony optimization (ACO)-based algorithm to solve the MT-TVRPTW-UES model, specifically designed to tackle large test instances that cannot be handled by CPLEX software. To demonstrate the viability and superiority of the MT-TVRPTW-UES model, we present two case studies based on real-world data from a large logistics company located in Hong Kong. The results show that the MT-TVRPTW-UES model significantly improves the MiC module demand satisfaction, environmental protection, and people’s social life.