Torc Robotics
Senior, ML Engineer - 3D Reconstruction
$199K–$299K
Remote - US, Ann Arbor, MI +1 · Remote · Senior · Full-time · Posted Nov 25, 2025
- Location
- Remote - US, Ann Arbor, MI
Remote - US - Experience
- Bachelor’s degree route: 6+ years; Master’s degree route: 3+ years; PhD route: 1+ years.
- Education
- Bachelor’s, Master’s, or PhD in Computer Science, Robotics, Electrical Engineering, or a related technical field.
About the job
Leads development of machine-learning-based 3D reconstruction, lane-line mapping, and pose-correction systems that produce annotations from logged autonomous-vehicle sensor data.
What you'll do
- Design, deploy, and improve offline reconstruction, lane-line detection, mapping, and pose-correction pipelines.
- Set technical direction, lead complex projects, and mentor engineers across lane-line, mapping, and pose-estimation work.
- Manage data and quality workflows, including dataset curation, map updates, and measurement of annotation and mapping quality.
What they ask for
- Strong hands-on experience with lane-line detection and creation, network reconstruction, and training lane-line models.
- Experience mapping previously unmapped areas and updating maps as road geometry changes.
- Strong pose-correction background, including diagnosing and correcting sensor calibration errors.
- Advanced experience in active learning, pseudo-labeling, computer vision, deep learning, and model training; plus depth in at least two listed areas such as detection, tracking, sensor fusion, segmentation, SLAM, or BEV.
Nice to have
- PPK/RTK GPS experience.
- GIS experience.
Pay
$199,200 - $298,800 USD base salary
Tools
Python, PyTorch, Lightning, Ray, MLflow, Weights and Biases, Parquet, Docker
AI focus
Develops and leads ML models and algorithms for lane-line detection and reconstruction, automated mapping, pseudo-labeling, and 3D reconstruction for autonomous driving.
Model-written summary · the original posting is authoritative