Type:
Status:
Goal:
The goal of this master thesis project is to conceive and develop a motion planning algorithm that is able to generate dynamic grasping trajectories for underactuated aerial systems. The dynamic grasping problem is completely different and much more challenging than its static counterpart. The motion-planning algorithm has to take into account the dynamic constraints of the system (e.g., underactuation, thruster saturations) and environment constraints (e.g., presence of obstacles). The planner has to generate motions that do not require the robot to stop while grasping. A major novelty of the thesis will be the tight combination of motion planning and control methodologies. The thesis will be divided in the following phases:
• Thorough modeling of the aerial grasping planning class of problems starting from [Spica et al, IROS 2012]
• Development of a novel approach to dynamic pick(&place) with aerial manipulators, possibly based on an extension of a geometric formulation using sampling-based algorithms [Simeon et al, IJRR 2004]
• Implementation and testing of the developed algorithm in a dynamic simulator
• Implementation of the method on the quadrotor –based aerial testbed of LAAS (optional)
Requirements:
• Study in the field of computer science, engineering, physics, or mathematics
• Strong background in planning algorithms, automatic control, and robotics
• Programming skills in C++ and Matlab
• Motivation to work in an interdisciplinary project and international environment
• Good English skills
Instructions:
In order to apply, please send an e-mail to af-applicants@laas.fr including:
1) your CV;
2) your Ms transcripts,
3) your Bs transcripts, and
4) the mail tag [mtp-PlanningAerialGrasping] in the subject of the e-mail
A complete description of the project is given in this flyer.
