In context
In 2024, social robotics was advancing toward human-robot collaboration in eldercare, healthcare, and industrial settings, yet most follow strategies positioned robots behind humans, which can cause psychological discomfort. Side-by-side accompanying was less explored but more natural and comfortable. This study addressed the need for motion control strategies that ensure comfort, compliance, and safety during human-accompanying tasks.
What was reported
Researchers from Fujian Agriculture and Forestry University and the Chinese Academy of Sciences proposed a control strategy based on admittance control for a two-wheel differential mobile robot to accompany a human side-by-side while avoiding obstacles. The method integrates an interaction force model derived from human-robot interaction space theory to prevent intrusion into the companion's intimate area, enhancing comfort. Admittance control parameters are tuned to improve motion compliance, and a behavioral dynamics model simulates human-like obstacle avoidance for safety.
Simulation results showed that the proposed method reduced robot velocity change by 69.6% compared to PID and 67.1% compared to virtual spring model (VSM), indicating superior compliance. The robot did not cause discomfort, and the obstacle-avoidance failure rate was only 10%, versus 40% for artificial potential field (APF) and 50% for VSM. In physical experiments, the robot achieved a failure rate of just 5%, with good compliance and comfort metrics.
Why it mattered
This work offered a practical framework for social robots to walk naturally alongside humans, addressing comfort and safety in dynamic environments. The demonstrated improvements in compliance and obstacle avoidance could accelerate adoption of companion robots in human-centric settings, such as factories where collaborative mobile robots operate near workers, enhancing human-robot interaction quality.
"The proposed method effectively achieves safe and friendly human accompanying and obstacle-avoidance control."
Source: 《机器人》期刊 (robot.sia.cn) · Published 2024-05-08 · “基于导纳控制的社交机器人伴随与避障控制策略”