Job offer description
Context :
In contrast to rigid robots, the kinematic behavior of deformable robot is not fixed during the design process. Any change (choice of material, addition of a sensor, addition of any other component, etc) could result in a change of the deformation behavior. Moreover, the relationship between the geometry of the robot and the deformation obtained may not necessarily be intuitive, and this is a major difficulty for a wider diffusion of deformable robotics.
The team DEFROST wish to develop tools to aid the design and optimisation process, and guide the user through the design and manufacture of deformable robots.
Objective :
The goal of this post-doc is to provide design assistance through optimisation methods and physics-based deformable models. The goal is to optimize the geometry, the material stiffness distribution, the placement of the actuators (…) in order to provide an adequate behavior of the robot. Reducing the computation time of such an optimization is also one of the objective, in order to allow the interaction with the user during the optimization. It is important that the user can easily improve and redefine the objectives of the optimization, given the results in a simulation.
Activities :
The post-doc will do the following activities :
• Conduct a bibliographical review on existing method on shape optimization for 3D printing and
deformable model simulation.
• Propose and implement an efficient method of optimization (using the framework SOFA and the plugin developed by the team)
• Improve the computation time in order to propose interactivity in the optimization.
More information and online application on the Inria website: On-line application
