MIT researchers, with support from Amazon, have developed a deep-learning model that utilizes a neural network architecture for warehouse robots. The model efficiently organizes robots into smaller groups, encoding their trajectories, origins, destinations, and relationships, enhancing coordination alongside traditional algorithms for thinning out traffic in a warehouse.

Read MIT News