Science Daily July 25, 2022 Multi-agent reinforcement learning (MARL) is a promising framework for solving complex tasks with many agents. However, a key challenge in MARL is defining private utility functions that ensure coordination when training decentralized agents. This challenge is especially prevalent in unstructured tasks with sparse rewards and many agents. Researchers at the University of Illinois have shown that successor features can help address this challenge by disentangling an individual agent’s impact on the global value function from that of all other agents. They used disentanglement to compactly represent private utilities that support stable training of decentralized agents […]
Tag Archives: and robotics
Small robot swimmers that heal themselves from damage
Science Daily March 24, 2021 Swimming robots that can “swim” through fluids and carry out useful functions, such as cleaning up the environment, delivering drugs, are often made of brittle polymers or soft hydrogels, which can easily crack or tear. Researchers at UC San Diego made swimmers that were 2 cm long in the shape of a fish that contained a conductive bottom layer; a rigid, hydrophobic middle layer; and an upper strip of aligned, strongly magnetic microparticles. Platinum in the tail, which reacted with hydrogen peroxide fuel to form oxygen bubbles propelled the robot. In a petri dish filled […]