Senra Systems · posted · checked
Senior Robotics Engineer, Robot Learning
C++CypressGoPythonPytorch$200k – $260kHybrid · United StatesSenior
Who can apply
Hybrid · United States
Location: Cypress, California, United States
At Senra Systems, we're building a new model for complex manufacturing, starting with electrical wire harnesses. We're a real manufacturer, not a lab: we build harnesses on our own production lines for some of the largest aerospace and defense companies in the country. Long term, automating this work is how we help bring complex manufacturing back to the U.S.
We're looking for a Senior Robotics Engineer to own the technical implementation of robot learning in our Automation Program. High-mix assembly is one of the hardest problems in manufacturing and has long resisted automation: deformable wire, small parts, and work that changes from job to job. Fixtured, programmed cells handle what they can. You'll take on the rest with learning-based methods, starting with tasks where parts vary from job to job, such as machine tending, and building toward dexterous wire manipulation. We buy robot arms, hands, and base models and build what is specific to harnesses: the data, the training, and the skills. Real hardware, on a real line, in real production.
This is a hands-on senior role on a small team with aggressive goals. You'll report to one of Senra's cofounders, our Head of AI & Robotics, who owns the technical vision for the program. You'll partner with the other senior engineers on the team to execute it, and help shape where it goes as we build. It's a lot of work, with a lot of ownership, and what you build goes onto the line fast.
You'll work closely with our Senior Robotics Engineer, Production to develop and deploy robotic capabilities on the factory floor. Your focus will be on robot learning, perception, and manipulation, using real production data to improve performance and address failure modes. The goal is to build systems that work reliably in production, not just demonstrate capabilities in the lab.
What you'll own:
Learned manipulation on our production line, from research to a skill running on real work, starting with machine tending: picking, loading, and feeding wire and parts that vary from job to job
Feeding what you learn back to production: learned skills, perception and grasping models, and failure analysis, handed to our Senior Robotics Engineer, Production to fold into cells on the line
The data pipeline for learned skills: demonstrations from teleoperation and station cameras, production data from our cells, labeling, and training and evaluation pipelines
Dexterous and wire-manipulation research on the tasks that have resisted automation: small parts, deformable wire, and contact-rich assembly