PIL 23 - Summer Internship
Problem Statement - Design and implement a multi-agent reinforcement learning (MARL) algorithm for task scheduling in a collaborative robotic system, with a focus on package transportation within warehouses. The goal is to improve efficiency and reduce the number of deadlocks, while operating in a non-deterministic environment.
Proposed Solution - We propose to use a variant of Q-learning, such as Independent or Cooperative Q-Learning, to train a team of robots to work together to transport packages. The robots will learn from their interactions with the environment and each other, and will be able to adapt to changes in the environment and task requirements. Deadlock avoidance will be incorporated into the learning process to reduce the number of deadlocks.