Home | deutsch  | Legals | KIT
Photo von Andreas Schmid

Dr.-Ing. Andreas Schmid

group: Steuerung und Regelung für Robotersysteme (SRR)
room: 004.1
phone: +49 721 608-44262
fax: +49 721 608-47141
andreas schmidUim2∂kit edu

Research Groups / Projects

'German Humanoid Robot Project' (SFB 588), Project 'Human-Robot-Cooperation' (K2)

The collaborative research center "German Humanoid Robot Project" aims at developing robot systems that are capable of performing everyday tasks in cooperation and multimodal interaction with human beings within a normal human environment. More

Research Fields

  • * Human-Robot Cooperation *

     

  • System Architecture

    The goal of this research area is to define a system architecture for the research center's complex robot sytem in collaboration with the other projects of the research center. The system architecture needs to integrate all system modules such as sensors, actors, sensor data fusion and processing, motion control, environment model, system control, and planner, as well as describe the data and control flow among the modules. This project K2 plays a central role in the design and implementation of the higher level modules like planner, system control, cognition, and resource administration.

  • Tactile Language

    The goal here is to develop a module that recognizes characters written on a tactile sensor and translates them into symbols. This allows for the definition of a tactile language to complement other human-machine-interfaces such as speech or gesture recognition. Since the research center's demonstrator robot will be clad almost entirely in an artificial skin there will be plenty of uses for such a tactile language, making robot communicaton and control as intuitive as possible.

  • Mobility

    The human-robot-cooperation methods developed so far are based on a static one-arm system. There cooperation methods are now to be extended such that they can be employed by a mobile robot system such as our research center's demonstrator robot. This extension requires a sensible partitioning of the movement between all axes (mobile platform, torso, arm, hand) of this highly redundant system. In doing so we aim to incorporate recorded trajectories of human motion and their characteristics to achieve human-like motion for our humanoid robot.

  • Safety

    The robot system of our research center is being designed to provide services to human beings in their everyday environment. Therefore conservative safety concepts as used for industrial robots (closed cages) are not applicable. The goal of this research area therefore is to split up the environment of the robot into zones of different safety levels. For each zone a corresponding bevahior pattern has to be defined along with a rule base that controls the transitions between different behaviors, depending on the safety level, sensor data and the environment model of the robot.

  • Arm-Hand-Coordination

    A coordination of arm (7 DOF) and hand (10 DOF) is required in order to grasp objects. It makes sense to swith the control of the whole system back and forth between arm and hand as necessary, depending on the phase of the grasp procedure. The next objective then is to extend this coordination using appropriate sensor information to allow for the grasping and touching of moving objects and human beings.

Miscellaneous