Related Vendors

FAULHABER Drive Systems

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

Wuxi InfiMotion Technology Co., Ltd.

Analysis Through Flexible Multibody System Modeling

During the project, the picking robot was modeled as a flexible multi-body system (FMKS) that considers both large rigid body movements through the driven axes and elastic deformations. Very precise laser measurements have shown that the main causes of inaccuracies at the TCP are particularly the tilting of the robot platform and the deflection of the telescopic rail structure. Both effects are influenced by the extension length of the telescopic rail because the moment acting on the robot increases with the extension length, causing greater deformations. Unlike a pure black-box modeling, the FMKS approach allows for a certain degree of modularity and adaptability if model parameters should change.

Data-Driven Approach Optimizes Picking Robot

For the preliminary parameterization and validation of the model, a data-driven approach was pursued. High-frequency camera recordings of the robot were created. A point-tracking algorithm implemented in Matlab was used to determine the deformation coordinates of the robot from the camera images. The data was then used to approximate the flexible multi-body system to the observed deformation through optimization methods. Despite a relatively small number of degrees of freedom, the flexible multi-body system sufficiently accurately represents the behavior of the real system. The resulting model was finally linked to the virtual commissioning platform ISG virtuos via the co-simulation interface Functional Mock-Up Interface (FMI). This makes it possible to derive a digital twin with elastic behavior. It is also used for model-based control concepts to compensate for elastic TCP deviations in real time.

Everything Stands and Falls with the Robot Control Concept

The flexibly modeled robot kinematics are also used for the drive control of the machine. The generated system models are used to optimize the machine dynamics through model-based control. A model with the lowest possible order is required for real-time calculation of the model-based control on the robot control system. Based on this model, a feedforward control for the cascaded control system is developed. This so-called 2-degree-of-freedom control aims to maximize the robot's performance while minimizing its susceptibility to vibrations. Additionally, the tilting of the TCP resulting from compliance is also considered and compensated in the control. An algorithm is developed to calculate the position deviation at the TCP. The result is then fed back to the control and compensated by the drive controller. The dynamics and accuracy of the system can be enhanced with the models to effectively counteract challenges from tilting and vibration problems due to the lightweight design.

This is how the Future of Lightweight Picking Robots will be

The developed methods are then tested with the picking robot. The model must be both sufficiently accurate and capable of real-time operation to be implemented into the control system. An experimental validation will also demonstrate that the digital twin with extended dynamic behavior can help minimize the robot's susceptibility to vibrations and improve positioning accuracy.

We thank the Ministry of Economic Affairs, Labor and Tourism Baden-Württemberg for the financial support of the project (BW1_4140).

*Valentin Leipe and Lukas Steinle (Institute for Control Engineering of Machine Tools and Manufacturing Units at the University of Stuttgart). Jonas Scheid and Marcel Hagedorn (both from robomotion GmbH). Philipp Rodegast and Joerg Fehr (Institute of Technical and Numerical Mechanics), and Denis Pfeifer (Institute of Technical and Numerical Mechanics and also ISG Industrial Control Technology GmbH).

Future-Proof Authentication with Universal RFID Readers

Universal RFID Readers: One Solution for All Credentials