DIANA

DYNAMIC & INTERACTIVE ROBOTIC ASSISTANT FOR NOVEL APPLICATIONS

KUKA INNOVATION AWARD 2016 PROPOSAL

RWTH Aachen in cooperation with Robots in Architecture – RWTH & RiA

Univ.-Prof. Dr. techn. Sigrid Brell-Cokcan, IP, Robots in Architecture
Univ.-Prof. Dr. rer. nat. Sabina Jeschke, IMA/ZLW & IfU, Verteilte Systeme
Johannes Braumann, Robots in Architecture, UFG Linz

The programming of industrial robots is commonly static, however in order to enable the use of robotics in new fields involving novel applications, these programs need to adapt to their surroundings. Due to their high complexity and unstructured nature, construction sites can be seen as an ideal testing environment of the universal applicability of robots for a number of reasons: Robot behavior needs to dynamically adjust to environmental constraints and construction tolerances. Lot sizes and repetitiveness of tasks are lower than with most production environments leading to low utilization of robotics. Most industrial robots are not transportable or mobile enough for usage at the construction site. Due to fluctuation and low degree of expertise within the workforce a number of people without a technical background need to be able to use, calibrate and re-program a robot for construction tasks.

We therefore propose a new approach towards robot programming. Combining research results from the areas of learning by demonstration with skill and parameter based robot programming we want to redefine the current teaching process and create a Dynamic and Interactive robotic Assistant for Novel Applications (DIANA). We demonstrate the capabilities of our approach within the assembly of building elements.

diana_concept