Refraction AI is on a mission to bring the future of last-mile delivery to life. With our fleet of REF-1 autonomous mobile delivery robots we are growing quickly, propelled by exciting advances in robotics technology, demographics, business models, social and regulatory policies. As consumer expectations and requirements change, so must the services and businesses that serve them. In a sector that is struggling to keep up due to legacy costs, dated infrastructure and poor integration, Refraction AI is a breath of fresh air, demonstrating what is possible with a little innovation, a dash of brilliance, and a whole lot of common sense.
- Improve the safety, robustness and reliability of the perception stack running on our robots, including tuning existing algorithms and developing new ones when necessary.
- Develop and train deep learning models to improve the quality of the robot’s understanding of its environment. Implement the models developed to run in real-time on embedded hardware onboard the robot.
- Work closely with the planning and control team to ensure smooth interaction with perception.
- Characterize capability and limitations of perception software stack. Forecast how future development work will adjust perception capabilities.
- Meet regularly with the operations team to understand the environment and conditions experienced by robots during deliveries. Use information from operations to guide development of the perception stack and improve overall system performance.
- Develop, test and release new features as part of the software team with agile sprints.
- Additional duties as assigned
- M.S. in Robotics, Computer Science or similar technical field
- Excellent background programming in C++ and Python
- Experience with deep learning frameworks (PyTorch, Tensorflow, etc.)
- Experience developing perception algorithms based on cameras, lidar, and/or radar
- Good working knowledge of Linux operating systems and build tools such as bazel or cmake
- Ability to work as part of a team
- Excellent communication skills, both written and verbal
- Excellent attention to detail
- Works well with uncertainty and ambiguity
- Successful completion of a background check
- PhD in Robotics, Computer Science or similar technical field
- 5+ years of work experience in related field
- Experience implementing algorithms on real robots
- Knowledge of robotics frameworks such as ROS, autoware, apollo cyber
- Experience with robotic simulation tools (e.g. CARLA or LGSVL) and developing unit tests
Reports to: Director of Software Engineering
Work Location: Remote, Ann Arbor, MI
Job Type: Full-time
Employment Opportunity and Benefits of Employment: We are committed to creating and fostering a diverse and inclusive workplace environment for all of our employees. We are an equal opportunity employer
This is a salaried, exempt position. We offer a competitive compensation package, excellent benefits and generous PTO. The company will provide the necessary technology resources.
Applying: It is strongly encouraged that you include links to any publicly accessible software repositories you have created or made significant contributions to (e.g. github, gitlab or bitbucket projects)