Research

Robots@IAS

Currently we have four robot platforms in our lab: PR2, TUM Rosie, iCub and Bender. The PR2 was provided by the Willow Garage Beta Program. TUM Rosie has been built around the KUKA-omnidirectional base and KUKA-lightweight LWR-4 arms. The iCub is a small-size humanoid robot being designed by the RobotCub Consortium, consisting of several European universities. B21 is the alumnus of robots in our robot-park.

Perception for Robots

While many specific perception tasks have been addressed in the context of robot manipulation, the problem of how to design and realize comprehensive and integrated robot perception systems for manipulation tasks has received little attention so far. We investigate a perception system that is tailored for personal robots performing pick-and-place tasks, such as setting the table, loading the dishwasher, and cleaning up, in human living environments.

Knowledge Processing

We investigate knowledge representation and processing mechanisms for mobile robots by grounding knowledge structures into the data structures and the perception and action mechanisms of the robot and by combining description logics knowledge bases with data mining, (self-) observation modules and imported knowledge from the World Wide Web. Our knowledge learning and processing capabilities include statistical relational learning and reasoning and naive physics prediction.

Plan-based Control

In the plan-based approach of Cogito, robots produce control actions by generating, maintaining, and executing plans that are tailored for the robots' respective tasks. The PARA project develops plan-based control mechanisms for human-robot interaction, where the robot assists the person in everyday tasks and adapts to the person's abilities, expectations and preferences.

Cognitive Manipulation

The CogMan project (1) develops computational and control models of pick-and-place tasks in the context of everyday manipulation activities in human environments, (2) implements the model into a control system for the kitchen scenario, and (3) empirically analyzes the impact of this control model on the flexibility, robustness, adaptability, and naturality of the robot behavior.

CRAM

CRAM equips autonomous robots with lightweight reasoning mechanisms that can infer control decisions rather than requiring the decisions to be pre-programmed. In this way, CRAM-programmed autonomous robots are much more flexible, reliable, and general than control programs that lack such cognitive capabilities. CRAM does not require the whole domain to be stated explicitly in an abstract knowledge base. Rather it grounds symbolic expressions from the knowledge representation into the perception and actuation routines and into the essential data structures of the control programs.

Perception of Human Activities

The automated perception of human activities is of great importance for any robot that has to co-exist or cooperate with humans in a social environment. Within this research field, the MeMoMan project aims at developing new computational models and a system for accurate measurement of human motion. Its primary goal is to develop markerless vision-based tracking algorithms for use with the industry-proven anthropometric human model RAMSIS, thus enabling new fields of application in ergonomic studies and industrial design while benefiting from the ergonomic expertise that has affected RAMSIS' design. The Multi Joint Vision project aims at creating a flexible and expandable system for distributed real-time computer vision applications, with application to human-robot interaction. Using a high density of optical sensors networked to multiple image processing nodes, the primary goal of the project is to enable a real-time interpretation of image data depicting human activities.

Facial Expression Recognition

The research project FaceMimic aims at enabling machines to utilize communication channels natural to human beings, such as gesture or facial expressions. The MuDiS project aims at granting machines the ability to adapt to typical human behavior. The goal of the project is the development of a multimodal dialog system that considers various human communication channels such as facial expressions, spoken language and gestures for human-machine interaction.

ASPOGAMO

ASPOGAMO stands for Automated SPOrt Game Analysis MOdel. The research project 'Sensor-based, Automatic Analysis of Football Games' is an ambitious, mid-term research project that studies the automation of these tasks. The main objectives of the project are (1) the investigation of novel computational mechanisms that enable computer systems to recognize intentional activities, (2) the development of an integrated software system to automate game interpretation and analysis, and (3) the demonstration of the impact of automated game analysis on application areas, such as sport science, football coaching, and sports entertainment.

Past Projects