Kinesthetic teaching improves robot precision and speed in industry
PAKT: Physically-Aligned Kinesthetic Teaching for Reinforcement Learning
Robotics
Summary
Robots often struggle to learn complex tasks that require very fine movements and speed in industrial settings. The paper presents a method called PAKT that lets humans guide robots physically by pushing or moving them, while making sure the robot only follows motions it can actually perform. This way, the robot learns from realistic demonstrations that respect its physical limits. The system uses a smart controller to turn low-speed actions into smooth, precise movements that handle contact well. Tests showed that PAKT helped robots finish assembly tasks faster and with fewer corrections.
What this means in practice
- •For industrial robot operators: Reduce cycle time and interventions when teaching robots complex contact-rich tasks in manufacturing settings.
- •For automation system integrators: Integrate kinesthetic teaching that respects robot physical limits to improve robot policy learning from human guidance.
Authors
Lars Johannsmeier, Yashraj Narang
Abstract
Real-world reinforcement learning (RL) systems still struggle with the demands of contact-rich industrial manipulation, including micrometer-level precision, success rates above 99%, and human-level cycle times. Although off-policy algorithms can improve performance by leveraging demonstrations and interventions, a key bottleneck is the lack of an intuitive interface for collecting such guidance while complying with constraints of the physical system and the policy. We propose PAKT, a framework for kinesthetic teaching in RL. As opposed to teleoperation approaches, PAKT relies on kinesthetic guidance, which is widely used in industry. However, a critical weakness of kinesthetic guidance is the possibility for the operator to move the robot along trajectories (e.g., velocities, accelerations, jerk) that the robot and/or policy cannot physically reproduce. Using PAKT, operators guide the robot through admittance control, which maps human-applied forces to motion. The downstream reference generator applies the same kinematic limits used during policy execution, keeping the collected trajectories within these limits. To support this teaching interface with an appropriate execution layer, PAKT adds a high-performance control stack that maps low-frequency RL actions to high-frequency torque commands. It consists of a reference generator and subsequent impedance controller, where the reference generator preserves the tracking performance of the impedance controller while improving contact handling and producing smoother policy actions. Across the reported runs on four insertion and industrial assembly benchmarks, including a data center compute tray, the end-to-end system reduces cycle time by 23%-48% and cumulative intervention count by 62%-86% relative to the HIL-SERL baseline. Project website: https://pakt-website.github.io/pakt-website}{https://pakt-website.github.io/pakt-website