Research Engineer, Human Biomechanics for Robot Learning

Socket.dev; San Francisco (CA); On-site; USD 90,000 - 150,000; Full time

Job summary

Relari in San Francisco is hiring an engineer-researcher who blends biomechanics with robot learning. You will design experiments and build data pipelines spanning EMG, force, tactile sensing, motion, and vision to create transferable representations for cross-embodiment manipulation.

Work closely with robot-learning researchers to validate representations on real hardware, influence hardware choices, and ensure data collection protocols meet biomechanics and downstream learning needs.

Qualifications

• Experience collecting and analyzing biomechanical or physiological data (EMG, force, pressure, tactile, motion).
• Strong skills in signal processing, statistics, time-series modeling, sensor fusion or ML.

Responsibilities

• Develop signal-processing and learning methods for synchronized multimodal data (EMG, force, tactile, pressure, motion, vision).
• Design experiments and sensor configurations to capture how humans regulate contact, stability, coordination and effort during manipulation.
• Build multimodal representations linking neuromuscular activity with external forces and touch feedback.
• Create data-collection protocols grounded in biomechanics and downstream robot-learning needs.
• Collaborate with robot-learning researchers to test representations on real robotic systems and assess cross-embodiment transfer.

Skills

Biomechanics

Signal processing

Python

Machine learning

Sensor fusion

Education

PhD welcome

Job description

Relari in San Francisco is hiring an engineer-researcher who blends biomechanics with robot learning. You will design experiments and build data pipelines spanning EMG, force, tactile sensing, motion, and vision to create transferable representations for cross-embodiment manipulation.

Work closely with robot-learning researchers to validate representations on real hardware, influence hardware choices, and ensure data collection protocols meet biomechanics and downstream learning needs.