Adaptive large neighborhood search for drayage routing problems involving longer combination vehicles

Longer combination vehicles (LCVs)—tractor-trailer combinations with two or more trailers—can move more cargo per trip than standard trucks, potentially reducing costs and emissions. Acknowledging this, the drayage industry has recently launched initiatives to increase the adoption of LCVs, aiming to mitigate supply chain issues. This paper aims to support these kinds of initiatives by tackling drayage routing problems involving LCVs and further aspects, such as heterogeneous truck fleets, containers of any size and cargo category, and compatibility constraints based on the containers’ sizes and loads. As solution method, we propose an adaptive large neighborhood search (ALNS) heuristic, whose search operators account for aspects such as truck/container compatibility and load configurations specific to each truck. The proposed ALNS heuristic also incorporates a novel acceptance criterion that seeks to escape local optima while allowing for revisiting the current best solution and avoiding search stagnation. Through a series of benchmark tests against a state-of-the-art exact approach, we show that the proposed ALNS heuristic can consistently find high-quality solutions for a wide range of instances while, in some cases, cutting runtimes from hours or even days to a few minutes. We also show that the proposed acceptance criterion enables improved performance. Finally, we use the proposed ALNS heuristic to derive managerial insights into the savings delivered by different types of LCVs in drayage operations. • We consider a drayage routing problem involving longer combination vehicles. • We use a state transition logic to model longer combination vehicles. • We develop a state-of-the-art adaptive large neighborhood search heuristic. • We propose the pulsed threshold acceptance criterion. • We estimate the savings that longer combination vehicles can deliver.