Company

Job Description

Context

Advanced Driver Assistance Systems (ADAS) bring improvements in driving comfort and road safety while continuously evolving toward higher levels of automated driving (AD). However, the complex and uncertain nature of traffic environments poses challenges in the development of such systems. Vehicles must make decisions based on imperfect or noisy information. To ensure safe, robust, and efficient trajectories, various sources of uncertainty must be incorporated into the decision-making process.

Lane changing is one of the key maneuvers in ADAS and AD since it enables vehicles to adapt to traffic conditions, overtake safely, navigate around obstacles, and follow desired routes. Making safe and reliable lane change decisions requires the vehicle to respect the dynamics constraints while interacting with other surrounding road users, whose behaviors cannot be predicted perfectly. Recent research therefore investigates methods to capture uncertainty before and during executing lane changes.

The internship is involved in an industrial research project, focusing on decision-making for lane changing maneuvers in dynamic and uncertain environments. The work aims to contribute to the development of intelligent decision-making algorithms capable of handling uncertainty efficiently, while assessing their potential for deployment in future vehicle systems.

Objectives

The primary goal of the internship aims to study various methods to make safe and efficient lane changes. Then, potential methods will be tested and evaluated on simulated real-world scenarios. Their performance will be analyzed with respect to risk of collision, travel time, and driving comfort. Finally, the implemented methods will be challenged by uncertain factors on the road such as the noisy measurements, and ambiguous behaviors of surrounding vehicles. Based on the results obtained, potential improvements will be identified and discussed for future development.

Opportunities

During the internship, the candidate will have the opportunity to:

• Gain foundational knowledge of ADAS and automated driving architectures.

• Develop solid understanding of lane changing maneuvers and their role in autonomous driving systems.

• Gain familiarity with uncertainty-aware decision-making and planning methods for driving applications.

• Work with simulation platforms and software development tools used in automotive research.

• Gain hands-on experience with algorithm design, implementation, and experimental validation.

• Contribute to Renault's industrial research activities addressing real-world autonomous driving challenges.

• Collaborate with experienced researchers and engineers in the field of intelligent mobility

Missions

During the internship, the candidate will

• Review state-of-the-art methods for lane changing maneuvers.
• Design a theoretical representation of a highway lane changing scenario.
• Implement and integrate potential decision-making algorithms for lane changing maneuvers within the simulation environment.
• Test simulated scenarios with uncertain factors that can influence the performance of lane changing.
• Analyze the performance of the methods with respect to safety, efficiency, comfort, and robustness to uncertainty.
• Document the results and present recommendations for future developments.

Your profile

• Master’s student (final year) in Computer Science, AI, Robotics, Mechatronics or a related field.
• Good understanding of vehicle motion planning and control.
• Basic understanding of algorithms, machine learning, or optimization.
• • Knowledge of sequential decision-making and reinforcement learning is a plus.
• Python and/or C++ programming
• Experience with simulation frameworks or automotive simulators is a plus.
• Strong analytical and problem-solving skills.
• Curiosity and willingness to explore new research topics.
• Ability to work autonomously while collaborating within a multidisciplinary team.

Job Family

Contract Duration