SAFE GRASPING OF TRANSPARENT OBJECTS

MOTIVATION

The developments of recent years in the field of artificial intelligence (AI) have helped robotics gain entry into areas far from automotive production and optimize their processes. These technological developments promise further scaling potential, including in the chemical and optical industries. However, they face particular challenges in automation, as they primarily work with glass bodies for which there is currently no flexible solution for industrial gripping.

GOAL AND APPROACH

The state of the art achieves a success rate of 72 percent for grasp point detection on glass bodies (ClearGrasp, Google, Columbia University, ICRA 2020). The MENTOR project aims to increase the success rate for grasping glasses and glass bodies to over 98 percent. To achieve this, a sensor head consisting of up to seven sensors working in different modalities will be developed, which will be integrated into various overall systems. These include a laboratory automation system based on articulated robots and a system for automatic testing of optical components.

INNOVATION AND PROSPECTS

The main challenges lie in recognizing the contours of transparent objects and their position in space. To master these and other challenges, the following innovations are sought in the overall framework: utilization of sensors in different wavelength ranges and modalities; data processing and fusion directly in the sensor head; creation of open interfaces for use in robotics and beyond. With this approach, a sensor system is to be developed that can be flexibly used beyond the application cases addressed in the joint project.

OUR CONTRIBUTION

Our contribution focuses on the development of a robotic modular system for handling glass containers in laboratory environments. Our tasks include not only the development of a data processing platform to enable added value services but also the definition and provision of standardized industrial interfaces for the sensor head to connect to the control level. In addition, we take on the evaluation of further data processing concepts for utilizing the data processing platform to provide added value services.