Dynamic coordination of intralogistics vehicle fleets with resource and routing constraints

To ensure the material flow of goods within warehousing and manufacturing, companies rely on intralogistics systems as an essential part of their supply chain management and production planning. Part of the material flows within intralogistics can be covered by intralogistics vehicles operating in a fleet to handle incoming transport orders. Vehicles can be operated either autonomously or manually, and vehicle fleets can be fully autonomous, fully manual or hybrid. The pandemic and the growth of e-commerce, combined with a shortage of labour and space, have put high pressure on intralogistics. To ensure a well-performing system with minimal down times, minimal risk of accidents, minimal energy use, and maximal throughput, proper coordination among vehicles in the intralogistics vehicle fleet is primordial. Automation of this coordination and optimising coordinating decisions could provide solutions in terms of performance, flexibility, and resilience of intralogistics vehicle fleets. To accommodate, this research addresses automated and optimised coordination of Multi-Robot Systems while focusing on high-level coordination and making abstraction of the vehicle type. Hence, the research results could be useful within coordination of both autonomous, manual, and hybrid intralogistics vehicle fleets. Relatedly, informal discussions with companies dealing with intralogistics vehicle fleets and a review of the literature on multi-robot coordination revealed some gaps in research and industry needs: (i) the literature on multi-robot coordination was scarce and outdated at the start of the research; (ii) optimised resource management (fuel, electrical energy, transport capacity, etc.) has received limited attention in the literature and in practise; (iii) the models used for travel between two locations in decentralised multi-robot task allocation are mainly based on Euclidean distance and rarely consider robots with limited resources- and routing constraints; (iv) the market puts new requirements for intralogistics vehicle systems, such as flexibility and openness. To accommodate, this research investigates automated and optimised coordination of Multi-Robot Systems by means of modular and decentralised dynamic coordination with optimised resource management while using more realistic models for travelling between two locations in the decentral multi-robot task allocation process. In contrast to static coordination, dynamic coordination makes coordination plans during task execution and can be done in real time. First, the state of the art on algorithms and techniques for multi-robot coordination is reviewed, with a focus on decentralised decision-making. Second, an optimised resource management approach for a single robot is proposed, determining the best refilling station and duration for replenishing resources. Furthermore, the effects of considering delays due to limited resources and traffic in the decentralised multi-robot task allocation process are investigated. Finally, a modular and decentralised coordination framework is proposed for the automated and optimised coordination of a Multi-Robot System.

Jaar van publicatie:2022

Toegankelijkheid:Open