ABSTRACT
This paper investigates the use of Q-learning for making optimal sequential product lifecycle decisions in sustainable manufacturing. Increasingly, sustainability is recognised as a key competitive strategy in manufacturing and considerations of sustainability dimensions is very important for optimal decisions in product lifecycle systems. Sustainable product lifecycle problems are usually sequential in nature with inherent uncertainties. These characterising attributes of sustainable product life cycle evolution systems inspires the use of reinforcement learning for sequential decision making in such systems. Q-learning is an established model-free reinforcement learning technique. Computational experiments reported in this paper show that Q-learning technique can help optimise sustainable product lifecycle decisions through each of its lifecycle phases.
