Open dataset captures human movement during retail robot guidance tasks
OHRID-Retail: An Open Multimodal Dataset of Human Activity in Retail Environments
RoboticsHuman-Computer Interaction
Summary
It can be hard for robots and people to work together in stores, where people move around, pick items, and interact with robots. The authors created a new open dataset by recording 16 people doing a shelf-picking task while a robot moved at different speeds nearby. They collected detailed body movement and muscle signals to show how people behave with and without robot guidance. This dataset helps improve technologies that recognize human activities and make robots better teammates in shared spaces like stores.
What this means in practice
- •For robot developers: Improve robot navigation and behavior by using human movement data during shared shelf picking tasks in retail settings.
- •For ergonomics consultants: Analyze muscle activity and movement patterns to design safer and more comfortable work routines for retail employees working with robots.
- •For fitness technology designers: Develop wearable systems that monitor body movement and muscle engagement by leveraging multimodal sensor data from real-world tasks like picking items under robot guidance.$Commercial implications: Creates advanced body movement tracking products for consumer or workplace use that integrate multi-sensor data such as inertial and EMG signals.
Authors
Xiangrui Wang, Yuetong Wu, Jalen Beeman, Robert Cook, Yu Gu, Nathanial Pearson, Trevor Smith, Read Hayes, Boyi Hu
Abstract
Open datasets describing human behavior in environments shared with mobile robots remain limited, particularly for retail activities that combine locomotion, reaching, object handling, and robot guided movement. This paper introduces OHRID Retail, an open, human centered multimodal dataset collected from 16 healthy adults performing a simulated shelf picking task under three within participant conditions: no robot, low speed robot guidance, and high speed robot guidance. Each participant completed two trials per condition. Whole body kinematics were recorded using 17 Xsens Awinda inertial sensors and muscle activity was measured at 10 locations using Delsys Trigno surface electromyography sensors. Descriptive analyses demonstrate variation in whole body movement intensity and muscle activation across robot interaction conditions and body locations. OHRID Retail provides openly available raw recordings, processed measures, documentation, and reproducible analysis resources. The dataset can support research in human activity recognition, multimodal sensor fusion, occupational biomechanics, ergonomics, human aware robot navigation, and human robot interaction in retail and related shared environments.