Research Intern, Scene Understanding and Planning

DiDi Labs • San Jose, CA

5d • Onsite

About The Position

DiDi Autonomous Driving, which became the independent company Voyager in 2019, develops Level 4 autonomous driving technology. The team aims to create a reliable, scalable autonomous fleet by combining DiDi’s data, operational expertise, and mobility ecosystem, with a focus on safety and user experience.

The Research Intern will help improve how the vehicle’s planner understands the current physical world, enabling more robust handling of complex real-world driving scenes without relying solely on static maps. The intern will tackle open problems in real-world driving using research methods and own work from problem definition through validation. Results may be published, with support for writing, while prioritizing solutions with product impact. The intern reports directly to the team lead and collaborates with research and engineering teams to move successful prototypes into the planning stack.

Requirements

• Be currently enrolled in an M.S. or Ph.D. program in Computer Science, Robotics, Electrical Engineering, Machine Learning, or a related field for the full duration of the internship.
• Be available for a full-time, onsite internship lasting 3–6 months; start dates are flexible.
• Strong programming skills in Python.
• Experience with modern deep learning frameworks such as PyTorch or JAX.
• Hands-on experience designing, training, and debugging deep learning models for computer vision, 3D perception, or sequential decision-making.
• Solid understanding of autonomous driving fundamentals, including perception, prediction, and motion planning.
• Ability to critically read research, form hypotheses, and conduct well-controlled experiments.
• Excellent communication and collaboration skills.

Nice To Haves

• Publications at leading venues such as CVPR, ICCV, ECCV, NeurIPS, ICLR, ICML, ICRA, IROS, CoRL, or IV.
• Experience with multimodal sensor fusion involving camera, lidar, or radar, or with BEV representations.
• Experience with transformers, graph neural networks, or other structured scene representations.
• Experience with uncertainty estimation or learning-based prediction or planning.
• Experience with large-scale model training and data pipelines.
• Experience with public autonomous driving datasets such as nuScenes, Argoverse 2, or the Waymo Open Dataset.
• Proficiency in C++; CUDA is a plus.
• Experience with robotics systems.

Responsibilities

• Survey recent research in scene understanding, prediction, and planning, and adapt it into practical approaches suited to system constraints.
• Design and implement deep learning models that produce richer, real-time understanding of the driving environment from onboard sensors and other available sources.
• Train and evaluate models on large-scale driving-log data and public benchmarks, and lead failure analysis to identify where and why models fall short.
• Define metrics and evaluation methods that connect perception quality to vehicle driving performance.
• Integrate successful prototypes with the planning stack in collaboration with production engineers.
• Collaborate with Perception, Prediction, Planning, and Mapping teams to move results from experiments into products.
• Work on research problems from definition through validated solution and, where appropriate, prepare results for publication.

What This Job Offers

• Job Type: Full-time
• Career Level: Intern