Multimodal Teleoperation Dataset for Fragile Object Manipulation with 221 Trials

221 trials from 11 novice participants manipulating objects with 20 visually encoded fragility levels (50–1000 gf). The dataset includes synchronized multimodal observations: robot joint trajectories, gripper force signals, multi-view RGB video, users' perceived fragility, confidence ratings, and trial outcomes. It was released by Jin Ong on figshare in April 2026.

Use Cases

• Modeling operators' internal estimates of object fragility based on pre-contact kinematics
• Analyzing human behavior patterns and failure modes in fragile object manipulation
• Developing personalized user modeling and behavioral profiling for teleoperation
• Designing proactive intervention strategies based on perceived fragility and confidence ratings
• Training adaptive robot control systems using synchronized multimodal observations

Strengths

• Includes 221 trials from 11 participants, providing a basis for behavioral analysis
• Captures synchronized multimodal data: robot trajectories, force signals, multi-view RGB video, and user perceptions
• Features 20 distinct fragility levels (50–1000 gf) to test a wide range of conditions

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• Data may reflect bias inherent to the experimental setup with novice participants

Provenance

Source

figshare

Collection Method

Experimental trials with novice participants teleoperating fragile object manipulation

Freshness

Last updated 2026-04-21 00:02:54