Autonomous Driving Intern: Semantic Interaction & Planning

Honda Ri; Mountain View (CA); Presencial; USD 34.000 - 48.000; Jornada completa

Descripción de la vacante

Honda Research Institute USA in Mountain View is seeking a research intern to develop semantic-aware interaction and behavior planning methods for autonomous driving. The intern will investigate how scene context, agent intent, interaction information, and semantic relationships can be used to make higher-level driving decisions such as when to proceed, yield, wait, reroute, or adapt to other road users.

The developed methods will integrate with HRI’s existing real-time navigation and

Formación

• Ph.D. candidate in Robotics, CS, EE, ME, or a related field.
• Strong background in behavior planning and motion planning.
• Experience in reinforcement learning or multi-agent systems.
• Experience developing and evaluating algorithms for robotics or autonomous systems.
• Programming proficiency in Python and/or C++.
• Familiarity with PyTorch or TensorFlow.
• Ability to conduct independent research and prototype solutions.

Responsabilidades

• Conduct research on semantic-aware behavior planning and interaction-aware decision making for autonomous driving.
• Develop methods that use scene context, agent intent, interaction relationships, and uncertainty to make high-level driving decisions.
• Investigate representations such as semantic features, scene/interaction graphs, multi-hypothesis agent behaviors, or learned behavior models.
• Develop decision-making strategies for interactions such as yielding, passing order, waiting, rerouting, and conflict resolution.
• Integrate developed methods with existing HRI navigation and motion-planning algorithms.
• Train and evaluate methods using simulation and real-world driving datasets.
• Evaluate both algorithm-level performance and impact on closed-loop autonomous-driving behavior.
• Collaborate with HRI researchers and contribute toward technical publications and system demonstrations.

Conocimientos

behavior planning

motion planning

decision making

reinforcement learning

autonomous driving

multi-agent systems

Educación

Ph.D. candidate in Robotics, Computer Science, Electrical Engineering, Mechanical Engineering, or a related field

Herramientas

PyTorch

TensorFlow

ROS/ROS2

IPOPT

CasADi

Descripción del empleo

Honda Research Institute USA in Mountain View is seeking a research intern to develop semantic-aware interaction and behavior planning methods for autonomous driving. The intern will investigate how scene context, agent intent, interaction information, and semantic relationships can be used to make higher-level driving decisions such as when to proceed, yield, wait, reroute, or adapt to other road users.

The developed methods will integrate with HRI’s existing real-time navigation and