Solving the online batching problem using deep reinforcement learning

In e-commerce markets, on time delivery is of great importance to customer satisfaction. this paper, we present a Deep Reinforcement Learning (DRL) approach for deciding how and when orders should be batched picked in warehouse minimize the number tardy orders. particular, technique facilitates making decisions whether an order individually (pick-by-order) or batch with other (pick-by-batch), if so which We problem by formulating it as semi-Markov decision process develop vector-based state representation that includes characteristics system. This allows us create deep reinforcement learning solution learns strategy interacting environment solve proximal policy optimization algorithm. evaluate performance proposed DRL comparing several batching sequencing heuristics different settings. The results show able produces consistent, good solutions performs better than heuristics.