Staff Robotics Software Engineer, AI/ML

General Motors•Warren, MI

1d•Onsite

About The Position

We are seeking a Staff Robotics Software Engineer, AI/ML to join the AI Research team in the Autonomous Robotics Center (ARC). In this role, you will lead robot learning and perception/grasp intelligence development for manufacturing manipulation systems, driving architectural decisions and guiding engineering work from prototype through production deployment. ARC is developing end-to-end robot policies that enable dexterous manipulation in real-world manufacturing environments. The team advances embodied AI by integrating multimodal perception, robot learning architectures, and physical robotic systems to solve manipulation, autonomy, and simulation challenges at industrial scale.

Requirements

M.S. or Ph.D. in Computer Science, Robotics, Electrical Engineering, Mechanical Engineering, or a related field.
7+ years of experience developing and deploying AI/ML systems for robotics or embodied intelligence applications.
Strong programming skills in Python and/or C++ with deep learning frameworks such as PyTorch or TensorFlow and production software engineering practices.
Experience with 3D perception systems including RGB-D, stereo, or LiDAR, along with calibration, coordinate frames, and real-world failure modes.
Demonstrated ability to lead technically without formal people management, including cross-functional alignment and technical execution.
Comfortable working in hardware-centric environments with iterative debugging and close integration with planning and control systems.

Nice To Haves

Ph.D. with publications in leading robotics or machine learning venues.
Experience with ROS 2 and MoveIt2 integration patterns for robotic manipulation systems.
Experience optimizing on-robot inference using tools such as TensorRT or ONNX Runtime.
Experience implementing dataset and experiment governance systems such as DVC or MLflow in robotics or machine learning environments.

Responsibilities

Lead development of perception and grasp intelligence capabilities, including segmentation, object and pose estimation, scene understanding, and grasp candidate generation and ranking.
Develop robot AI models such as diffusion policies, ACT, VLM/VLA-guided agents, and imitation learning approaches for manipulation and autonomous task sequencing.
Design and scale training-to-inference pipelines with strong evaluation practices, reproducibility, and runtime performance requirements including latency, determinism where required, and robustness to domain shifts.
Define model-to-system interfaces including inputs/outputs, uncertainty signaling, failure modes, and fallback strategies in coordination with planning and controls.
Drive measurable system outcomes and work with validation teams to translate failures into prioritized data and model improvements.
Provide hands-on technical leadership through architecture design, code reviews, documentation, and mentorship for engineers implementing model pipelines and robotics integrations.
Contribute to intellectual property, publications, and external collaborations when they accelerate capability development.