OpenRoboCare: Expert Caregiving Demonstrations with 19.8 Hours of Multimodal Data

A multimodal dataset capturing 19.8 hours of expert demonstrations across 315 sessions. It includes synchronized RGB-D video, tactile sensing, eye-gaze tracking, pose annotations, and action labels from 21 occupational therapists performing 15 daily caregiving tasks. The dataset was contributed by the EmPRISE Lab at Cornell University and is hosted on AWS Open Data.

Use Cases

• Training robot learning from demonstration models based on temporal action annotations and pose tracking.
• Developing multimodal perception systems for robots based on synchronized RGB-D video, depth maps, and tactile readings.
• Studying safe human-robot interaction protocols for caregiving based on expert demonstrations and first/third-person videos.
• Analyzing expert task execution and ergonomics based on eye-gaze tracking and pose data.

Strengths

• Includes 19.8 hours of data across 315 expert demonstration sessions.
• Captures data from 21 domain experts (occupational therapists) performing 15 distinct tasks.
• Integrates multiple synchronized modalities: RGB-D video, tactile sensing, eye-gaze tracking, pose, and action labels.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

EmPRISE Lab at Cornell University

Collection Method

Recordings of expert occupational therapists performing daily caregiving tasks.

Time Range

null

Freshness

null

Geography

null