Senior Machine Learning Perception Engineer - Fallback Driving System
Build and optimize ML perception models (object detection, tracking, prediction) for GM’s fallback autonomy stack using camera, lidar, and radar data to safely bring vehicles to a stop when primary autonomy fails.
- Compensation
- $170.6K–$261.3K / yr
- Location
- GM Automation - Sunnyvale - GM Automation - Sunnyvale
- Work setup
- Not specified
- Seniority
- Senior
- Employment
- Not specified
- Experience
- 3+ years
- Category
- Ml / AI
- Posted
Skills
Must have
- BS, MS, or PhD in Machine Learning, Robotics, Computer Science, or a related technical field; or equivalent practical experience building ML perception systems.
- 3–5 years of experience developing ML solutions in perception, prediction, and/or autonomous driving or related domains.
- Strong experience with multi‑modal sensor data (camera, lidar, radar), including data preprocessing, synchronization, and fusion.
- Deep expertise in modern deep learning for perception, such as convolutional and transformer‑based architectures for:
- 2D/3D object detection
- Semantic and instance segmentation
- Multi‑object tracking and motion prediction
- Proficiency in at least one major ML framework (e.g., PyTorch, TensorFlow, JAX) and Python for model development, training, and analysis.
- Solid software engineering skills, including experience working in C++ or similar languages in large, collaborative codebases.
- Demonstrated ability to define ML metrics, design experiments, and systematically improve model performance and robustness.
- Strong problem‑solving, communication, and cross‑functional collaboration skills.
- Self‑motivated, with a passion for autonomous driving technology and its potential impact on safety and mobility.
Nice to have
- Experience deploying ML models on embedded or resource‑constrained platforms, including model optimization and performance tuning for real‑time inference.
- Experience with AV/ADAS perception stacks, robotics, or ROS.
- Familiarity with safety‑critical systems and development practices.
- Experience with large‑scale data pipelines, labeling workflows, and experiment management for ML.
Full description
Job Description
At General Motors, our product teams are redefining mobility. Through a human-centered design process, we create vehicles and experiences that are designed not just to be seen, but to be felt. We’re turning today’s impossible into tomorrow’s standard —from breakthrough hardware and battery systems to intuitive design, intelligent software, and next-generation safety and entertainment features.
Every day, our products move millions of people as we aim to make driving safer, smarter, and more connected, shaping the future of transportation on a global scale.
As a Senior Machine Learning Engineer on the State Estimation and Mapping (SEAM) organization, you will develop and improve the ML perception model that powers the secondary (fallback) autonomy stack for Super Cruise 3. You will focus on building robust perception from multi‑modal camera, lidar, and radar data so the vehicle can safely bring itself to a stop when the primary autonomy stack is unavailable.
You will lead the design, implementation, and continuous improvement of ML models for object detection, segmentation, tracking, and prediction, working closely with partner teams across perception, planning, controls, and safety.
What You'll Do
- Design, train, and evaluate ML perception models for object detection, semantic/instance segmentation, tracking, and short‑horizon prediction using multi‑modal camera, lidar, and radar data.
- Develop and maintain the secondary stack perception model that enables the fallback autonomy system to safely bring the vehicle to a minimal risk condition when the primary system experiences a fault.
- Define clear ML success metrics (e.g., precision/recall, latency, robustness under edge cases) and drive systematic experimentation to improve model performance against those metrics.
- Analyze large‑scale datasets, curate challenging scenarios, and build data selection and labeling strategies that improve robustness for long‑tail and degraded‑sensor conditions.
- Implement efficient training and inference pipelines, including model optimization techniques (e.g., pruning, quantization, distillation) to meet on‑vehicle compute and latency budgets.
- Collaborate with software and infra engineers to integrate models into production systems, including interfaces, configuration, deployment, monitoring, and regression safeguards.
- Partner with Safety, Systems Engineering, and Product to translate system requirements into concrete ML model requirements, metrics, and validation criteria.
- Contribute to verification and validation strategies for the fallback perception model, including offline evaluation, simulation, hardware‑in‑the‑loop, and on‑road testing.