Mixed Reality Robot Teach-in Assistant

Michael Spitzer*, Inge Gsellmann, Matthias Hebenstreit, Manfred Rosenberger

*Corresponding author for this work

Research output: Contribution to conferencePaper

Abstract

In the semiconductor industry, robots’ movements must be very accurate to meet the industry’s high requirements. In wafer manufacturing processes, robots need frequent calibration (teach-in) to ensure precise robot motions. However, the teach-in process is a complex and time-consuming task that requires highly trained technical experts with appropriate knowledge. As a result, corresponding personnel requirements at every production site, or rather high travel frequency, are necessary. Due to COVID-19 pandemic regulations, the last two years in particular have shown the impact of travel restrictions on the industry. An intuitive strategy for optimizing such teach-in processes must counteract this complex calibration task and save personnel resources and traveling costs.

The new interaction strategy described in this paper lets the engineer position a virtual robot on the HoloLens 2 by dragging it to a target and automatically computing the joint angles required to reach this position. By configuring the virtual robot model, the assistant forces the robot to stay in a predefined axis or prevent collisions, increasing safety for humans and machines. This verified position can then be sent to the real robot and used in the teach-in process. Additionally, the assistant supports creating learning and training scenarios for new employees. They could train the positioning without using the real robot by just moving the virtual model.
Building a lab environment mimicking real wafer sorter robots makes it possible to test new interaction strategies for viability immediately. The assistant was developed, tested, and verified in this lab environment. The next step will be implementing the lab scenario in a real industry scenario at semiconductor industry partners.

The Mixed Reality Robot Teach-in Assistant showed high potential to support the Robot Teach-in process and reduce the training time for new or not trained employees.
Original languageEnglish
Number of pages6
DOIs
Publication statusPublished - 2022
Event12th Conference on Learning Factories: CLF2022 - Singapur, Hybrider Event, Singapore
Duration: 11 Apr 202213 Apr 2022

Conference

Conference12th Conference on Learning Factories
Abbreviated titleCLF2022
Country/TerritorySingapore
CityHybrider Event
Period11/04/2213/04/22

Fingerprint

Dive into the research topics of 'Mixed Reality Robot Teach-in Assistant'. Together they form a unique fingerprint.
  • Mixed Reality Robot Teach-in Assistant

    Michael Spitzer (Speaker), Inge Gsellmann (Speaker), Matthias Hebenstreit (Speaker) & Manfred ROSENBERGER (Speaker)

    11 Apr 202213 Apr 2022

    Activity: Talk or presentationTalk at conference or symposiumScience to science

Cite this