Multi-Agent Reinforcement Learning for Value Co-Creation of Collaborative Transportation Management (CTM)

Abstract

Collaborative Transportation Management (CTM) is a collaboration model in transportation area. The use of CTM in today's business process is to create efficiency in transportation planning and execution processes. However, previous research paid little attention to demonstrate the ability for all agents in CTM to co-create value in services. The purpose of this paper is to increase the understanding of value co-creation in CTM area and learning processes in real systems based on value co-creation of CTM. Multiple case studies were used to analyze the value that was perceived by all agents in CTM in each collaboration stage and provided empirical evidence on the interactions among agents. Model-free reinforcement learning was used to predict how CTM could reduce transportation cost, increase visibility, and improve agility. The simulation results show that the input, feedback, and the experience of the agents are used to structure the collaboration processes and determine the strategies.