The work in this position will be carried out within the SFB “EASE – Everyday Activity Science and Engineering” (funded by the DFG), which is to perform fundamental research to enable robots to do human everyday activities, such as setting a table or loading a dish washer.
One strand of research of the CGVR lab in this endeavor is sub-project R03 (“Embodied simulation-enabled reasoning”), where we are to develop algorithms that enable robots to “envision” the outcomes of their tentative plans by simulating the real world around them before they actually follow through with their plan. This requires novel methods for faster-than-realtime physically-based simulation and the development of predicates in order to evaluate the outcomes.
In another strand, sub-project H01 (“Acquiring activity models by situating people in virtual environments”), we will deal specifically with the challenge of grasping. Here, the task is to develop novel methods such that a human user can grasp virtual objects using a virtual hand with high simulation fidelity and high dexterity. The goal is to allow humans to interact with virtual objects just like in real live, and to study their grasping behaviors under a large variety of scenarios and parameters.
The work will involve research on geometric real-time algorithms and physically- based simulation. In addition, it will involve computer graphics and virtual reality, in general. Also, you should be comfortable with applying mathematical methods, in particular, optimization methods and linear algebra (you don't have to prove theorems).
The successful candidate is expected to work towards achieving his/her PhD with this project. Your tasks will also comprise a small amount of help with teaching in the computer science program (e.g., as teaching assistant).
Qualifications:
Candidates should have an excellent Master’s degree in computer science, or a related discipline such as computational engineering, physics, mathematics, etc. You should be passionate about your work and highly self-motivated. Required skills are solid experience in C++ software development, and a very good command of English (reading/writing/speaking). Ideally, you have specialized in real-time computer graphics or physically-based simulation, you are capable of effectively applying mathematical methods, and you have knowledge in GPGPU programming. In addition, the successful candidate will be highly self-motivated, passionate about their work, and have good ability to work both independently as well as in a team in a multidisciplinary environment. Speaking German is a plus, but not required.
Further details:
PhD Positions, University of Bremen, Germany (2018)