About this role
Summary: Design, train, and deploy PyTorch models and C++ post-processing for online mapping in autonomous vehicles. The role involves creating 3D maps from sensor data, optimizing for real-time performance on NVIDIA DRIVE hardware, and ensuring safety and accuracy across L2-L4 systems.
Self-driving cars need to understand the road ahead in real time, even without a prior map. Our team builds the online mapping system that turns raw data from multiple sensors into an accurate, continuously updated 3D model of the driving environment. This model captures lanes, road boundaries, crosswalks, intersection layouts, and traffic control information to support decisions across the autonomous driving system. You’ll write the models, write the C++, and ship both into vehicles on the road, running on NVIDIA DRIVE Orin and Thor. The tasks you perform will be tested at scale, on real cars, with real safety requirements. NVIDIA’s autonomous driving stack spans the full range of automation, and online mapping is a core component across it, so your work will reach L2++, L3, and L4 systems. The role spans two subject areas, and we want someone who enjoys both:
- Learned mapping involves training DNNs to predict map elements directly from visual sensors, laser scanning, and radar. This includes owning the GT story and collaborating with partners to provide the vital data.
- Geometry and C++ engineering: building the post-processing that turns network outputs into stable, consistent, production-ready map elements.
What you’ll be doing:
- Design, train, and improve PyTorch models that predict online map elements (lanes, road boundaries, crosswalks, road markings, topology) from camera, LiDAR, and/or radar, using BEV and vectorized-map representations.
- Solve the hardest scenes: model intersection topology (connectivity between incoming and outgoing lanes, turn paths, merges and splits) and traffic light understanding (detection, state, and association of each light to the lanes it controls).
- Write production C++ for geometry-based post-processing: curve and surface fitting, association, multi-frame fusion and tracking, topology reasoning, and consistency checks on DNN outputs.
- Apply multi-view geometry, calibration, coordinate transforms, and sensor fusion to produce 3D maps that stay accurate and stable over time.
- Optimize and deploy models alongside algorithms on NVIDIA DRIVE Orin and Thor under real-time and safety constraints, and get them running in vehicles.
- Lead the narrative around GT and the data loop by planning datasets and defining GT and labeling requirements. Work with labeling, data, and tooling partners to produce high-quality GT. Mine data, choose metrics, and analyze failure cases to drive improvements.
- Define critical metrics and build evaluation tooling to measure quality gains against large-scale real-world data.
- Partner with perception, localization, planning, and infrastructure teams to integrate online mapping into the full AV stack, across L2++, L3, and L4 programs.
- Contribute to technical direction and architecture decisions across the learned and geometric components of the mapping system.
What we need to see:
- MS or PhD or equivalent experience in Computer Science, Electrical Engineering, Robotics, or a related field, or equivalent experience
- 5+ years of industry or research experience in autonomous driving, robotics, or computer vision.
- Strong foundation in 3D geometry: camera models, projective geometry, coordinate transforms, optimization, and curve/surface fitting.
- Strong C++ skills, with a track record of designing and shipping production-quality algorithm code that is fast, reliable, and maintainable.
- Hands-on experience training deep neural networks in PyTorch, including model design, training pipelines, debugging, and performance analysis.
- Experience with data-driven development: building datasets, defining labels and GT, choosing metrics, and improving models by analyzing failures.
- Excellent problem-solving and communication skills, and the ability to drive technical direction across teams.
Ways to stand out from the crowd:
- Direct experience with DNN-based online or HD map generation, or BEV perception (e.g., vectorized map prediction, lane topology learning), or equivalent experience to BEV perception.
- Experience with intersection understanding, such as lane connectivity and topology reasoning in sophisticated junctions, or traffic light detection and lane association.
- Experience combining learned models with geometric post-processing in a production AV system and publications or open-source contributions in 3D vision, mapping, or autonomous driving.
- Background in sensor fusion, multi-object tracking, or temporal fusion for map elements.
- Experience deploying DNNs and algorithms on embedded GPUs (CUDA, TensorRT), ideally on DRIVE Orin or Thor, under real-time constraints.
#AutonomousVehicles
Your base salary will be determined based on your location, experience, and other objective job-related factors. The expected base salary ranges for this position are:
184,000 USD – 287,500 USD for Level 4; and 224,000 USD – 356,500 USD for Level 5.
You will also be eligible for equity and [benefits](https://www.nvidia.com/en-us/benefits/).
Applications for this job will be accepted at least until November 13, 2026.
This posting is for an existing vacancy.
NVIDIA uses AI tools in its recruiting processes.
NVIDIA is committed to fostering an inclusive work environment and proud to be an equal opportunity employer. As we highly value diversity in our current and future employees, we do not discriminate (including in our hiring and promotion practices) on the basis of race, religion, color, national origin, gender, gender expression, sexual orientation, age, marital status, veteran status, disability status or any other characteristic protected by law.