- Company Name
- Toyota Research Institute
- Job Title
- Automated Driving Advanced Development Intern, Machine Learning Research
- Job Description
-
**Job Title**
Automated Driving Advanced Development Intern – Machine Learning Research
**Role Summary**
Conduct research, design, prototype, evaluate and integrate end‑to‑end machine‑learning models that convert raw sensor data into vehicle trajectories for autonomous driving. Work on scalable architectures, simulation and real‑world testing, and collaborate with researchers, data engineers, and autonomy engineers to transition models from prototype to production.
**Expectations**
- Deliver robust ML components for perception, planning, and control.
- Execute closed‑loop evaluations in simulation and real‑world tests.
- Iterate on multimodal, language‑conditioned models to improve generalization.
- Contribute to research publications and technical documentation.
**Key Responsibilities**
- Conduct research on generative modeling for vision‑based end‑to‑end planning.
- Implement scalable architectures handling raw sensor inputs (video, lidar, radar, IMU).
- Prototype, validate, and refine imitation‑learning and large‑scale data‑driven models.
- Perform closed‑loop evaluation in simulation environments (e.g., CARLA, DriveSim).
- Explore multimodal and language‑conditioned frameworks, transfer learning, and foundation models.
- Collaborate with ML engineering on data sampling, preprocessing, training, ablation studies, and deployment.
- Debug, profile, and optimize models on CUDA/NVIDIA stack.
- Maintain version control, testing protocols, experiment tracking, and ML‑ops pipelines.
**Required Skills**
- PhD candidate or equivalent experience in Computer Science, Robotics, Engineering, or related field.
- Proficient in Python; experience with PyTorch.
- Knowledge of version control (Git), testing, and software engineering fundamentals.
- Strong collaborative mindset for deploying reliable ML systems.
**Bonus Skills**
- ML engineering workflow: data sampling, curation, preprocessing, training, ablation studies, evaluation, deployment, inference optimization.
- CUDA debugging and profiling.
- Tools: Weights & Biases, MLflow, Metaflow.
- Experience with ROS, simulation frameworks, and vehicle interfaces.
- Knowledge of 3D perception, object detection architectures, and multimodal transformers.
- Experience with motion planning (trajectory optimization, sampling‑based planning, MPC).
**Required Education & Certifications**
- Current PhD or equivalent experience in Computer Science, Robotics, Engineering, or related discipline.
---