Main focus of research at the Autonomous Systems Technology Institute is the development of autonomous mobile robot platforms. For example, such systems are to be enabled to explore and navigate in unknown unstructured environments own their own. As a demonstration platform, MuCAR-3, a modified VW Touareg, is available, that is equipped with computer controlled actuation elements and sensors.
MuCAR-3: Munich Cognitive Autonomous Robot Car, third generation.
Research on perceptual capabilities hereby focusses own an agile multi-focal camera platform as the main sensor, supported by a high-resolution 360 degrees rotating laser scanner mounted onto the vehicle's roof. Together, both sensors allow algorithmic perception of traversable ground as well detection and recognition of objects relevant to the current mission.
MarVEye: Multifocal active/reactive Vehicle Eye, properties similar to that of the human eye (large field-of-view, high resolution, fast saccades)
Spatio-temporal models of all detected objects are used to build up an internal representation of the environment, based on which cognitive functions, lilke situation analysis or path planning, take place. Here, the own vehicle serves as the reference system so that independence from global references like GPS is gained as far as possible.
Some of the methods developped at our Institute were already successfully demonstrated at the DARPA Urban Challenge 2007 as well as at C-ElRob 2007 and Elrob 2008. Further, some of our methods were integrated into various military research projects, including visual lane keeping on country roads and marked lanes, LIDAR based traversability analysis and object tracking.