HABIT: Human-Aware Behavior and Interaction Training Dataset for Robot Manipulation
HABIT: Human-Aware Behavior and Interaction Training Dataset for Robot Manipulation
This research presents HABIT, a large-scale demonstration dataset for robot manipulation in human-present environments. The dataset includes over 10,000 episodes spanning more than 160 hours across 60 tasks, organized into three interaction roles: Collaborator (joint task completion), Coworker (separate tasks in shared space), and Supervisor (human directing the robot). Experiments indicate that training on this data enables robots to exhibit human-aware behaviors such as spatiotemporal synchronization, yielding, and gesture grounding, which are not produced by models trained solely on human-absent data. The work aims to extend robot policies to environments shared with humans by treating human presence as a new axis of dataset diversity.
The paper introduces HABIT, a dataset of over 10K episodes and 160 hours across 60 tasks designed to train robot policies for human-present environments, demonstrating improved spatiotemporal synchronization, yielding behavior, and gesture grounding compared to human-absent data.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
episodes
>10,000 episodes
Reported trials
View original evidence
dataset for human-present environments. We organize tasks into three roles capturing distinct modes of human-robot interaction: Collaborator, where human and robot jointly accomplish a task; Coworker, where they pursue separate tasks in a shared space; and Supervisor, where the human directs the robot. The dataset comprises over 10K episodes and over 160 hours across 60 tasks. Our experiments show that training on h
Open source S3Reported duration
>160 hours
View original evidence
dataset for human-present environments. We organize tasks into three roles capturing distinct modes of human-robot interaction: Collaborator, where human and robot jointly accomplish a task; Coworker, where they pursue separate tasks in a shared space; and Supervisor, where the human directs the robot. The dataset comprises over 10K episodes and over 160 hours across 60 tasks. Our experiments show that training on h
Open source S3tasks
60
View original evidence
dataset for human-present environments. We organize tasks into three roles capturing distinct modes of human-robot interaction: Collaborator, where human and robot jointly accomplish a task; Coworker, where they pursue separate tasks in a shared space; and Supervisor, where the human directs the robot. The dataset comprises over 10K episodes and over 160 hours across 60 tasks. Our experiments show that training on h
Open source S3
- paper: not_reported
What remains unknown
Dataset licensing and commercial-use rights are not established here.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
dataset for human-present environments. We organize tasks into three roles capturing distinct modes of human-robot interaction: Collaborator, where human and robot jointly accomplish a task; Coworker, where they pursue separate tasks in a shared space; and Supervisor, where the human directs the robot. The dataset comprises over 10K episodes and over 160 hours across 60 tasks. Our experiments show that training on h
Open source S3
arXiv:2606.31682v2 Announce Type: replace Abstract: Large-scale demonstration datasets have been central to recent progress in general-purpose robot policies. However, existing datasets are collected in human-absent settings, and policies trained on such data may perform tasks competently in isolation but fail to exhibit human-aware behaviors. To address this gap, we introduce HABIT, a large-scale robot demonstration
Open source S2
uman-present data elicits human-aware behaviors that robot-only data fails to produce: spatiotemporal synchronization in Collaborator tasks, yielding in Coworker tasks, and gesture grounding in Supervisor tasks. Moreover, training on HABIT enables rapid adaptation to new human-robot interaction tasks. By introducing human presence as a new axis of dataset diversity, HABIT extends robot policies to environments shared
Open source S4
with humans.
Open source S5
HABIT: Human-Aware Behavior and Interaction Training Dataset for Robot Manipulation
Open source S1
Implications for data suppliers
RoboSignal interpretation and collection questions, not statements of buyer demand.
- Confirm the required data type and collection setting with the buyer; this source does not establish a complete collection specification.
- Compare the reported units and scope before using these quantities in a budget. Recording hours, sensor-hours and trajectories are different measures.
- Validate demand and acceptance criteria with a buyer before scaling. Publication, popularity and a research result do not establish a purchase commitment.
Source：arXiv Robotics — research abstracts · arxiv.org