We are looking for a motivated Perception Engineer to join the Stretch Perception Team. In this role, you will be a key contributor to our perception stack, focusing on how our robots see, interpret, and interact with their surroundings. You will work on production level code that is already in the operations of our customers, helping us refine how our robots handle the unpredictability of the real world.
Our team provides robots with robust environmental awareness, covering core components like object detection, state estimation, and sensor fusion. You will apply a diverse toolkit of techniques ranging from geometric processing and filtering to AI/ML to ensure our robots operate reliably in messy, human centric environments. We are a hands-on team, in which you will regularly test on physical hardware, dive deep into logs to debug performance issues, and implement the fixes that make our products better.
In this role, you will
Implement and refine perception algorithms to solve real world sensing challenges, ensuring the robot accurately models its environment
Debug and troubleshoot complex perception issues, using data driven approaches and log analysis to identify root causes and improve system reliability
Write clean, efficient, and maintainable C++ code designed to run in real time on resource constrained hardware
Integrate multimodal sensor data like LiDAR, RGB-D, and IMU to build robust representations of objects and workspace state
Develop and maintain infrastructure, including test suites and visualization tools, to validate and iterate on your perception modules
Analyze field data and performance metrics to identify edge cases and drive continuous improvement of our models and algorithms
We are looking for
Education: BS or MS in Computer Science, Robotics, or a related field
Experience: 2+ years of professional experience with a BS or a recently completed MS with relevant project or internship experience
Software Engineering: Proficiency in modern C++ and a solid understanding of software development best practices like version control, unit testing, and code reviews
Perception Fundamentals: A strong grasp of core robotics concepts like state estimation, image processing, neural networks, stereo vision, and 3D data analysis
Problem Solver: A knack for debugging. You should be comfortable digging into sensor streams and telemetry to figure out why a robot’s world model does not match reality
Hands on Attitude: Experience working with physical hardware or sensors. You enjoy seeing your code move a robot in the real world
Collaborative Spirit: Ability to clearly explain technical trade-offs and work closely with navigation and manipulation teams to deliver a cohesive product