[Comp-neuro] PhD in interdisciplinary ML / Robotics / BCI project

Michael Tangermann ms at we-are.de
Thu Oct 30 00:53:53 CET 2014


Dear colleagues,
there is a new offer for an interdisciplinary Machine Learning / BCI / Robotics 
PhD position in Freiburg. Please circulate it in your lab and email lists.

Thanks a lot, Michael

--- snip ---

Doctoral Research Position at the University of Freiburg, Germany

The University of Freiburg invites candidates to apply for a fully funded 
three-year PhD position (TV-L E13). The job offers a novel perspective to 
relevant problems in robotics in an inter-disciplinary approach involving 
engineering, artificial intelligence / machine learning, neuroscience and 
assistive technology.

The interaction between a (semi-autonomous) robotic device (e.g., a robotic 
wheelchair) and a human user is often ruled by a set of strategies executed by 
the robot. Typical robotic tasks are to accomplish a navigation task, to map the 
environment or to manipulate objects of the environment. The user may either 
control the actions of the robot or be an interaction partner for the robot. The 
level of control the user has over the robot's actions can be placed on the 
continuum between absolute low-level control (e.g., a joystick), over mid-level 
control ("turn right on next occasion") up to high-level control ("take me to 
the kitchen"). While state-of-the-art robotic control already takes the changing 
environment into account before strategies are selected and executed, the 
internal state of the user and fluctuations thereof typically are disregarded. 
Examples of informative mental states of users would be the level of 
task-induced workload, the sense of control/agency in the current situation or 
the perceived complexity of the current situation.

In the project, relevant internal mental states of the user shall continuously 
be assessed. Therefore, electrophysiological recordings (e.g., EEG, MEG, GSR, 
HRV, etc.) will be exploited with methods developed in the field of 
brain-computer interface (BCI) systems. Upon successful decoding of a mental 
state of interest, the robotic behavior can be adapted specifically to an 
individual user and to the perceived situation. Using the resulting co-adaptive 
strategy between human and robot, we expect an enhancement of usability and user 
acceptance during various tasks in human-robot interaction. During the project, 
expected tasks will comprise offline experiments with assessment of 
electrophysiological signals, simulated robotic control, offline and online data 
analysis with BCI methods, online experiments with a robotic application and 
mental state monitoring, evaluation of adaptive strategies of the robot. The 
candidate is expected to actively participate in the dissemination of research 
results in publications, on conferences and by teaching.

The target group are candidates with a degree at master level in robotics, 
computer science, artificial intelligence / machine learning, electrical 
engineering or cognitive science, with a strong relation to science, technology 
or innovation studies. Candidates are expected to have practical experience in 
all of the three fields robotics, machine learning/AI, and Brain-Computer 
Interfaces. We furthermore expect outstanding qualifications, the motivation to 
obtain a Doctoral degree and the interest to work in an interdisciplinary 
environment. Finally we require applicants to have excellent communication 
skills in English.

Applications should be send to ais-applications at informatik.uni-freiburg.de. The 
application documents should consist of a single PDF file and include

- Motivation letter (Expose, respectively)
- Two letters of recommendation from professors or persons authorized
  to supervise dissertation projects
- Curriculum vitae
- Copies of the following documents:
  a. Higher education entrance qualification (school-leaving certificate)
  b. University examination certificates (intermediate examinations)
  c. University diploma
  d. Any documents providing evidence of academic achievements,
     relevant practical experience, and qualifications earned at or
     outside of the university

Further information can be obtained from Wolfram Burgard and Michael Tangermann 
({burgard|tangerm}@informatik.uni-freiburg.de).



More information about the Comp-neuro mailing list