Data-Video Robotics Generalist

AI Trainer Jobs; United States; Remote; USD 28,000 - 36,000; Full time

Job summary

AuraOne seeks a Data-Video Robotics Generalist to remotely review robotics outputs, evaluating perception, motion, and safety. You will flag failures and document corrections to help train models while ensuring safe operation in physical-world scenarios.

As a contractor, you’ll work with task-based volume and program pacing, applying structured rubrics to rating trajectories and perception labels, and communicating clearly about failure modes and standards.

Qualifications

• Hands-on experience with robotic, autonomous, or teleoperated systems.
• Familiarity with structured incident taxonomies and severity scoring.
• Clear written reasoning that names the failure mode and the standard being applied.
• Reliable async availability for at least 10 hours per week.

Responsibilities

• Review robotics review outputs against current safety standards and operating envelopes for Data-Video Robotics Generalist assignments.
• Grade perception, motion, and decision logic on a structured rubric.
• Flag near-miss and unsafe trajectories with severity tags that match incident taxonomies.
• Capture the corrected behavior or reference trajectory so the modeling team can train on it.

Skills

Robotics evaluation

Perception review

Motion safety

Incident taxonomy

Robotics review

ai annotation tools

attention to detail

english communication

remote self-management

video annotation

Tools

Simulation environments

Trajectory replay tooling

Job description

Data-Video Robotics Generalist is a remote review track for evaluating robotics review outputs against real-world physical-AI constraints.

Category: Robotics & Embodied AI · Pay: $23 / hr · Location: Remote — US-eligible · Contractor

About the role

Data-Video Robotics Generalist is a remote review track for evaluating robotics review outputs against real-world physical-AI constraints. Reviewers grade trajectory choices, perception assumptions, and safety envelopes; flag failure modes; and document the corrected behavior so the modeling team can train on it.
Robotics review systems run in the physical world, where a confident-looking failure can hurt someone. AuraOne uses experienced robotics and operations reviewers to grade outputs against perception, motion, and safety constraints.
Review teleoperation, navigation, and manipulation work. Physical-world safety is part of it.

Responsibilities

• Review robotics review outputs against current safety standards and operating envelopes for Data-Video Robotics Generalist assignments.
• Grade perception, motion, and decision logic on a structured rubric.
• Flag near-miss and unsafe trajectories with severity tags that match incident taxonomies.
• Capture the corrected behavior or reference trajectory so the modeling team can train on it.

What you should bring

• Hands-on experience with robotic, autonomous, or teleoperated systems. Operating, supervising, or evaluating them for Data-Video Robotics Generalist work.
• Comfort reasoning about perception, motion planning, and physical safety envelopes.
• Familiarity with structured incident taxonomies and severity scoring.
• Clear written reasoning that names the failure mode and the standard being applied.
• Reliable async availability for at least 10 hours per week.

Role signals

Example tasks

• Grade a robotics review model's chosen trajectory and flag any near-miss or unsafe motion.
• Score a perception output against ground-truth labels and rate the failure category.
• Adjudicate a disputed safety call between two reviewers using firm operating procedures.
• Audit a 25-trajectory batch for rubric consistency and report drift to the program lead.

Useful experience

• Background in robotics engineering, autonomy operations, or industrial safety.
• Prior experience reviewing AI-assisted robotics tooling and its failure modes.
• Familiarity with simulation environments and trajectory-replay tooling.

Compensation and schedule

$23 / hr
Expected arrangement: contractor , with program-defined task volume and review pacing. Placement depends on current program demand and reviewer confirmation.

Skills used in matching

• Robotics evaluation
• Perception review
• Motion safety
• Incident taxonomy
• Robotics review
• ai annotation tools
• attention to detail
• english communication
• remote self-management
• video annotation