Positive Friction

Positive friction, contrary to the traditional goal of frictionless systems, explores the beneficial use of carefully introduced resistance in various applications. Current research focuses on leveraging positive friction to improve decision-making in human-AI interaction, enhance robot locomotion on uneven terrain, and optimize control in manufacturing processes. This involves employing diverse methods, including reinforcement learning, large vision language models, and recurrent neural networks to model and manage friction, leading to improved efficiency, robustness, and safety in diverse fields. The impact of this research spans from enhancing human-computer interaction to advancing robotics and manufacturing technologies.