Engineer

Wandercraft Team ExpansionOur control team is growing to take on new challenges! In 2025, Wandercraft is venturing into humanoids, in close collaboration with the Renault Group. We are leveraging our expertise in bipedal robotics and expanding our skills in manipulation and load carrying, in a competitive industrial environment.We are therefore creating a manipulation team with a mission: to develop the knowledge and codebase to meet the many client use cases we receive, involving the dextrous manipulation of various parts.Robotics is undergoing a revolution with the advent of Machine Learning, enabling fluid movements and innovative, more natural choices of contact, on the ground as well as in interaction with objects. Nevertheless, traditional model-based approaches have their advantages in terms of interpretability and speed of iterations, and are still widely used in the industry as well as in our robots.In this team, we will explore new manipulation technologies (Machine Learning, imitation Learning, design iterations in benchmark of hands, demonstration techniques) with a pragmatic approach and keeping in mind the integrability with the rest of the software and hardware. We will sometimes use more traditional methods for simplicity or as a complement, for the first use cases. We will also try to prototype and compare different end-effectors (grippers, shelf hands, in-house designed hands etc.).The missions may include (non-exhaustive list):Literature reviews in particular in manipulationPresentation to the team of the latest innovations from the literature or the latest algorithms developed in-houseDevelopment of manipulation algorithms (or policy training)Tests on real and simulated robotsImprovements to the realism of the simulator, comparison of simulatorsAutomated tests, code quality, design of verification pipelines to ensure non-regressionPrototyping of end-effectors in collaboration with the Hardware team, prototyping of demonstration toolsDemonstrations of tasks for learning trialsOn-site visits to our clients' factories to ensure the proper functioning of use cases and correct as neededProfile SoughtDegree: Engineering degreeExperience: 5 years in the robotics sector, ideally with experience in manipulation.Academic experience will be valued, particularly in optimal control, imitation learning, or diffusion models.Ideally, you have experience in maintaining data over time (storage of datasets, corrections, augmentations, ease of use, tools, CI,...).Creativity, innovation, you are able to develop new things with a low existing base and make prototypes (personal or professional robotics projects highly appreciated).You have an attraction for the experimental and the adaptation of algorithms to the constraints of the terrain and a commercial product.You have experience in team software development (modularity, test-driven development, continuous integration,...).You easily manipulate elements of linear algebra (matrix calculations), kinematics and physical modeling (changes of reference frames, velocity calculations,...), and are experienced in software development (python/C++).Rigorous, you know how to reread and explain your calculations or algorithms clearly.Curious and with a good analytical mind, you are able to ensure effective technological surveillance and learn from scientific literature or meetings at conferences.SkillsPython: expertC++: autonomousMachine Learning (data analysis, augmentation, labeling, validation): expertImitation Learning, diffusion models: expertRobotics (kinematics, dynamics, real data analysis): autonomousLinear algebra, matrix calculations, numerical optimization: autonomousAutomatics (observers, controllers): autonomousVision (feature detection, map creation, SLAM): basic knowledgeGood software development practices, continuous integration, automated tests: autonomousAdditional InformationType of contract:Location:Level of education:Experience:Postuler