Exoskeleton Controls Engineer

Skip•San Francisco, CA

2d•Hybrid

About The Position

Skip is an early-stage, 28-person startup spun out of Google X in 2023, developing wearable robotic products to restore mobility and enable new movement experiences. The Exoskeleton Controls Engineer will design, develop, test, and validate the high-level control system for Skip’s powered leg exoskeleton, MO/GO. The role covers how the device senses wearer movement and provides natural, safe assistance across activities such as walking, stairs, and hills. This begins as an individual-contributor position with leadership for critical systems. The engineer will work closely with firmware, mechatronics, and product teams and contribute directly to product development. The position is full-time and hybrid, based at Skip’s office in San Francisco’s Dogpatch neighborhood, with a requirement to work at the Bay Area office more than three days per week. The engineer will also wear prototypes several hours a week for data collection, on-body testing, and feedback.

Requirements

At least 5 years of experience developing control systems for robotics, exoskeletons, prosthetics, or other physical systems that interact with people.
Strong foundation in control theory, state estimation, and sensor fusion, such as Kalman filtering.
Experience taking controllers from simulation to real hardware, including tuning and debugging physical systems.
Proficiency in C/C++ and Python, with experience deploying control code on real-time or embedded platforms.
Experience designing and running experiments with human subjects or real-world users and analyzing the resulting data.
Familiarity with biomechanics and human gait.
Experience designing systems at scale, with a focus on testing and validation.
Attention to detail.
Ability to work at the Skip Bay Area office more than three days per week.
Sense of humor and tolerance of Australian and Canadian spelling.

Nice To Haves

Experience with lower-limb exoskeletons or powered prostheses.
Experience with machine-learning-based or reinforcement-learning controllers deployed on real robots.
Familiarity with series elastic actuators and torque-controlled drives.
Experience with powered consumer electronics, such as drones or robot vacuums.
Experience in startup environments.
Experience using AI coding tools to broaden the impact of your skills.
Personal motivation to improve human movement.

Responsibilities

Own the high-level exoskeleton control architecture, from sensing and state estimation through assistance torque commands to the low-level motor controller.
Develop and tune gait-phase estimation, activity recognition, and transition detection for activities such as level walking, stairs, slopes, and sit-to-stand, using IMU, encoder, and torque-sensor data.
Design assistance strategies and torque profiles that adapt to users, speeds, and terrain, including human-in-the-loop optimization.
Explore and deploy learning-based control methods when they outperform classical approaches, and make them work reliably on embedded hardware.
Build and maintain simulation and modeling tools, such as MuJoCo, OpenSim, and MATLAB/Simulink, for controller development and validation.
Define and own safety behaviors, fault handling, and edge-case responses so the device fails gracefully on the body.
Plan and conduct human-subject testing and data collection, then use the resulting data to improve controllers.
Guide design and development decisions for future product iterations and Movewear systems, including sensor selection, actuator requirements, and compute needs.
Wear prototypes several hours a week for data collection, on-body testing, and feedback.
Contribute directly to product development, work with firmware, mechatronics, and product teams, and participate in team events.