Adaption of Construction Robots and Building Components for Enhanced Robustness in Contact Situations

Research Project 35-1 of the Cluster of Excellence Integrative Computational Design and Construction for Architecture -
EXC IntCDC RP35-1

This project aims to take the first steps towards our long-term vision of developing a robust robotic construction system for automated assembly and disassembly based on construction robot capabilities. Such a robotic construction system shall comprise a robot/robots, building components and automated assembly processes, which are all compatible to each other. This entire robotic construction system needs to be robust, reliable and safe. From a control engineering perspective, each parameter must be optimized so that the robotic construction system is as robust as possible against uncertainties.

We will define and investigate the respective parameters and requirements for such a robotic construction system. The requirements may regard the geometry, the material and friction, the joints, and the connection mechanism of the building elements, the precision of the robot, the motion capabilities of the robot, the sensors, the assembly strategy and the overall robustness of the robotic construction system.

Since it is especially challenging to determine and control the state of the robot and the process in contact situations, we will specifically study these situations to gain insights into the requirements for such a robotic construction system. We will investigate the dynamics during manipulation, interaction, and contact situations to explore the capabilities and limitations of robots for the assembly of components to develop successful, sensitive, and seamless control strategies for the assembly of building elements. We will describe and consider the interaction dynamics including kinematic chains and contact forces and their effects during robot motion planning and at each step from task scheduling to trajectory generation and joint control. We will explore control strategies for building component assembly using, for example, force control, reinforcement learning or data-driven control.

More information

Contact

This image shows Anja Lauer

Anja Lauer

Jun.-Prof. Dr.-Ing.

Tenure-Track-Professor

To the top of the page