Skip navigation

University of Rostock / Institute for Computer Science / Chair of Software Engineering (Head Prof. Dr.-Ing. habil. P. Forbrig)

Project Staff:

  • Mr Alexandru Bundea
  • Mr Muhammad Salman Tayyab


Tasks and research content:

In this subproject, the human-robot dialog is technically implemented. For this purpose, the training requirements, general psychological aspects and "neurological interaction aspects" are realized with the help of apps on tablets and the robot. For this purpose, specific psychological aspects that should positively influence the robot-patient "relationship" are stored algorithmically and rule-based. Furthermore, the data resulting from the analysis of the neurological aspects will be processed for evaluation by the robot.


For this purpose, training specifics, psychological and neurological interaction specifics are mapped in an integrated way ("script") in a domain-specific language ("Corporate Task Language, CoTaL") and thus allow a digital implementation. Based on this, automatic program code generation for the behavior of the robot during a therapy session can be enabled in the future without technical expertise.

The technique of "personas" is used to model these role profiles. Activities and characteristics are identified for the future users. An abstract dialog structure is derived from the analysis of the tasks to be performed and the objects used. As a result, state machines are applicable. Refinements to the concrete user interface are modeled manually or derived from relationship structures; for this purpose, "usability tests" are performed with test persons (healthy persons, persons with stroke). Parallel to the clinical validation, iterative system optimizations are then carried out in this subproject.

  • Portion of the therapy manuals for the various forms of therapy to be transferred to digital applications.
  • Use of the analysis of the peculiarities in the behavior of stroke patients and therapists for the particular selected forms of therapy.
  • Implementation of recommendations for the interaction of the humanoid robot in order to be able to react adequately to these "neurological interaction aspects" of the patients.
  • Digitalization of training tasks.
  • Evaluation of the usability of the developed applications.
  • Support of the partners from psychology and health care management in their evaluations (quality of the therapeutic "relationship" on one hand and the patient's preference evaluation when using the humanoid robot on the other hand).



Universität Rostock
Institut für Informatik
Lehrstuhl Softwaretechnik
Albert-Einstein-Str. 21
18059 Rostock


Telefon: 0381 498 7620
Email: peter.forbriguni-rostockde