(M/W) AI & Control Intern – Perception and Navigation

To serve new industrial use cases ever more reliably, our humanoid robot must be able to detect multiple objects of interest and estimate their poses. It must also be able to understand its environment in order to navigate safely and autonomously. 

Our team is responsible for developing the knowledge and the codebase needed to address these challenges. We explore a range of state-of-the-art techniques, whether based on machine learning, classical computer vision, or model-based control. 

The Perception and Navigation team is part of the AI & Control department and works closely with other members of this department. It also collaborates regularly with other R&D teams: embedded software, electronics, and mechanics. 

Your responsibilities may include: 

• Conducting literature reviews related to ongoing projects 

• Presenting the latest innovations from the literature or internally developed algorithms to the team 

• Developing algorithms for object detection, pose estimation, or feature recognition 

• Developing control algorithms for various robots, 

• Working closely with other teams to ensure the overall performance of the algorithms 

• Testing on real robots and in simulation 

• Traveling to customer sites (e.g., healthcare centers or factories) to ensure algorithms function properly in real-world conditions and making corrections as needed 

• Automated testing, code quality, and designing verification pipelines to ensure non-regression

We are only offering 6-month internships. Applications for shorter durations will not be considered.

Qualifications: You are in your last year of Masters or Engineering degree in perception, navigation, machine learning, robotics, or a similar field.

Experience: Hands-on experience in robotics, computer vision, machine learning or navigation - whether in industrial, academic or personal projects - will be valued.

Ideally, you are familiar with software development (Python/C++), particularly in a team (good practice, version control, etc.).

You are creative and innovative. You are comfortable in developing new things with a limited existing base and making prototypes. You enjoy experimental work and adapting algorithms to the constraints of real-world environments and commercial products.

You are comfortable with general algorithmic, and in particular with linear algebra (matrix calculations), and physical modeling (reference frame changes, odometry, etc.). You are rigorous, able to proofread your work, and explain your calculations or algorithms clearly.

You are curious and have good analytical skills, enabling you to monitor technological developments effectively and learn from scientific literature or meetings at conferences.

Skills (hard and softskills) :

• Python - Autonomous level

• C++ - Autonomous level

• Vision (feature detection, mapping, SLAM...) - Autonomous level

• Navigation (path planning, path tracking...) - Basic knowledge

• Machine learning (data analysis, augmentation, labeling, validation) - Basic knowledge

• Robotics (kinematics, odometry, real data analysis) - Basic knowledge

• Linear algebra, numerical optimization - Autonomous level

• Control systems (observers, controllers) - Basic knowledge

• Continuous analysis of the state of the art in the field - Basic knowledge

• Software development best practices, continuous integration, automated testing - Basic knowledge