Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
HEAP Whole Body Motion Planning
In this project we would like to enable our autonomous excavator (http://www.rsl.ethz.ch/robots-media/heap.html) to plan a motion such that it reaches the desired end-effector pose.
The use case for this project is final end-effector pose adjustment before executing the tree harvesting maneuver, large scale manipulation and landscaping. The planning framework [2] has successfully been applied to tracked robots [1] and a mobile base with robotic arm and we aim to scale it up in this project. The strength of the proposed approach is that it takes into account the full dynamics of the system and therefore enables execution of more complex motions. Demonstrating the algorithm’s performance on a large system with many degrees of freedom systems such as our excavator would be a big contribution towards autonomous construction and autonomous forestry.
[1] Giftthaler, Markus, et al. "Efficient kinematic planning for mobile manipulators with non-holonomic constraints using optimal control." Robotics and Automation (ICRA), 2017 IEEE International Conference on. IEEE, 2017.
[2] Farshidian, Farbod, et al. "Sequential linear quadratic optimal control for nonlinear switched systems." IFAC-PapersOnLine50.1 (2017): 1463-1469.
The use case for this project is final end-effector pose adjustment before executing the tree harvesting maneuver, large scale manipulation and landscaping. The planning framework [2] has successfully been applied to tracked robots [1] and a mobile base with robotic arm and we aim to scale it up in this project. The strength of the proposed approach is that it takes into account the full dynamics of the system and therefore enables execution of more complex motions. Demonstrating the algorithm’s performance on a large system with many degrees of freedom systems such as our excavator would be a big contribution towards autonomous construction and autonomous forestry.
[1] Giftthaler, Markus, et al. "Efficient kinematic planning for mobile manipulators with non-holonomic constraints using optimal control." Robotics and Automation (ICRA), 2017 IEEE International Conference on. IEEE, 2017.
[2] Farshidian, Farbod, et al. "Sequential linear quadratic optimal control for nonlinear switched systems." IFAC-PapersOnLine50.1 (2017): 1463-1469.
- Familiarizing with the existing planning algorithm and integrating it into simulation
- Implementing the tracking controller and/or running the planning in the receding horizon fashion
- Hardware tests
- Bonus: incorporate actuator uncertainty
- Bonus: implement the self-collision avoidance constraints
- Familiarizing with the existing planning algorithm and integrating it into simulation - Implementing the tracking controller and/or running the planning in the receding horizon fashion - Hardware tests - Bonus: incorporate actuator uncertainty - Bonus: implement the self-collision avoidance constraints
- Independent working style and self-motivation
- Excellent programming skills C++
- Optimization knowledge & control knowledge
- ROS knowledge
- Independent working style and self-motivation - Excellent programming skills C++ - Optimization knowledge & control knowledge - ROS knowledge
Please contact Dominic Jud (dominic (dot) jud (at) mavt (dot) ethz (dot) ch) and Edo Jelavic (edo (dot) jelavic (at) mavt (dot) ethz (dot) ch) for any questions. Your application should include a very brief motivational statement, your transcript of records and your CV.
Please contact Dominic Jud (dominic (dot) jud (at) mavt (dot) ethz (dot) ch) and Edo Jelavic (edo (dot) jelavic (at) mavt (dot) ethz (dot) ch) for any questions. Your application should include a very brief motivational statement, your transcript of records and your CV.