You will be part of the Robotics Research Group at the Department of Mechanical Engineering, Division of Production engineering, Machine design and Automation (PMA). The group has pioneered robotics research in Europe since the mid-1970s and was among the first to develop active force feedback for assembly operations. Already in 1980 it developed learning insertion algorithms based on stochastic automata. It has covered virtually all aspects of sensor-based robotics, from the high-level task specification down to low-level sensor-based control, and applied the research results in a variety of industrial applications. In the last decade the group shifted its attention towards service robots (behaviour-based mobile manipulation, shared control, learning control), medical robotics (natural interfaces, haptic bilateral control), industrial robot assistants, and active sensing. PMA has created several spin-off companies that are active in robotics-related activities, has initiated several free and open-source software projects in robotics (Orocos, KDL, iTaSC, eTaSL, …), and has participated in a large number of EU projects in robotics, mostly oriented towards control and software development, with a focus on model-driven engineering techniques. More information is available through the link below.
Next to the robot and peripheral (sensors, fences, grippers, safety equipment …) hardware, the programming effort is a main cost driver in industrial robotic applications.
Typical robots are programmed using a robot-specific language via an offline development environment, in combination with point teaching on the robot system itself (guiding the robot to key poses and storing this information). This way of working requires skilled robot programmers, and also makes robots only viable for larger production volumes with limited variability (factory workers can make no or only small adjustments to the robot program).
In order to make robots more flexibly deployable, a lot of research and development is being conducted in making more high level and intuitive robot programming interfaces (cf. for example the latest products of UR, KUKA and Franka Emika). A step further is programming (or learning) by demonstration. In this paradigm, a robot task is inferred from one or a set of demonstrations.
Within the robotics research group at KU Leuven, we are focussing on different aspects and applications of programming by demonstration, and are looking for two excellent PhD candidates. Both positions will build upon and extend the approaches given in the references below. A main concept in our approach is to combine imitation learning with a constraint-based model-based approach. In this way, we can focus learning on aspects that are too uncertain or too much varying to model, yet avoid learning aspects of the task that can be easily modelled. As a result, the required learning time/number of demonstrations can be reduced significantly compared to a purely data-driven approach, and more formal guarantees on the robot behaviour during task execution can be given.
For a first PhD position, the goal is to develop automatic learning techniques to derive robot skill parameters and sequences of subtasks. For example, when performing assembly tasks, a number of parameters related to interaction force/torque thresholds must be determined, and it would overload the operator to require him to specify all these values manually for each variation in the task.
In a second PhD position, the goal is to investigate generalizing task demonstrations to different workpiece geometric properties by decomposing demonstrations to work-piece dependent and work-piece independent aspects.
 Vergara C., De Schutter J., Aertbeliën E., (2019). Combining Imitation Learning With Constraint-Based Task Specification and Control. IEEE Robotics and Automation Letters
 Vochten M., De Laet T., De Schutter J. (2018). Robust Optimization-based Calculation of Invariant Trajectory Representations for Point and Rigid-body Motion. In: 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (5598-5605). Presented at the 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Madrid, 01 Oct 2018-05 Oct 2018. doi: 10.1109/IROS.2018.8593540.
 Vergara C., Van de Perre G., El Makrini I., Van Acker BB., Saldien J., Pintelon L., Chemweno P., Weuts R., Moons K., Burggraeve S., Vanderborght B., Aertbeliën E., Decré W. (2018). Improving productivity and worker conditions in assembly Part 2: rapid deployment of learnable robot skills. Presented at the IROS2018 Workshop on Robotic Co-workers 4.0: Human Safety and Comfort in Human-Robot Interactive Social Environments, Madrid, Spain, 01 Oct 2018-05 Oct 2018.
 Aertbeliën E., De Schutter J. (2014). Etasl/eTC: A Constraint-Based Task Specification Language and Robot Controller Using Expression Graphs. In: 2014 IEEE/RSJ International Conference on Intelligent Robots and Systems (1540-1546). Presented at the IEEE/RSJ International Conference on Intelligent Robots and Systems, Chicago, 14 Sep 2014-18 Sep 2014.
 Arbo MH., Pane Y., Aertbeliën E., Decré W. (2018). A System Architecture for Constraint-Based Robotic Assembly with CAD Information. In: IEEE International Conference on Automation Science and Engineering (CASE): [proceedings]. IEEE Conference on Automation Science and Engineering Presented at the 14th IEEE International Conference on Automation Science and Engineering, Technical University of Munich, Campus Garching, 20 Aug 2018-24 Aug 2018.
 De Schutter J., De Laet T., Rutgeerts J., Decré W., Smits R., Aertbeliën E., Claes K., Bruyninckx H. (2007). Constraint-based task specification and estimation for sensor-based robot systems in the presence of geometric uncertainty. International Journal of Robotics Research, 26 (5), 433-455.
A successful candidate has obtained a MSc degree in engineering (Mechanical, Mechatronics, Electrical, Computer Science) related to Robotics and has a background and interest to contribute to:
Contributions to free and open source software projects (also beyond the topic of the project!) and hands-on experience with robot platforms and sensor systems (vision, force …) are both a plus. If applicable, please list them clearly in your application or send us your portfolio.
In your motivation letter or extended CV description, please consider to mention your previous experiences and skills, which may help to make relevant contributions to the project.
The selected candidate is furthermore expected to:
The successful candidate will receive:
A start in the course of 2019 or first quarter of 2020 is to be agreed upon.
Please use the online application tool to submit your application.
- an academic CV with photo
- a pdf of your diplomas and transcript of course work and grades
- statement of research interests and career goals (max. 2 pages)
- sample of technical writing (publication or thesis)
- contact details of at least two referees
Note: the position might be filled in earlier if an excellent candidate is found.
For more information, send an e-mail to firstname.lastname@example.org.
You can apply for this job no later than October 15, 2019 via the online application tool
KU Leuven seeks to foster an environment where all talents can flourish, regardless of gender, age, cultural background, nationality or impairments. If you have any questions relating to accessibility or support, please contact us at diversiteit.HR@kuleuven.be.Continue reading
|Title||PhD Position on Robot Programming by Demonstration|
|Job location||Oude Markt 13, 3000 Leuven|
|Published||August 7, 2019|
|Application deadline||October 15, 2019|
|Job types||PhD  |
|Fields||Artificial Intelligence,   Computer and Society,   Computer Communications (Networks),   Human-computer Interaction,   Programming Languages,   Computer Engineering,   Robotics,   Mechanical Engineering,   Mechanics,    and 1 more. Electronics  |