AUDIo – Audit solution for ML-based, data-driven services
Table of contents
Important data at a glance
| Project titel: | AUDIo – Audit solution for ML-based, data-driven services |
|---|---|
| Subproject title: | DBB measuring system |
| Duration: | 09/2019 – 08/2022 |
| Collaborative project: | yes |
| Funding: | Förderprogramm: „Smarte Datenwirtschaft“ |
| Funding code: | 01MD19005B |
| Processor / Contact person: | |
|
Cooperation: |
|
| Logline: | Als Projektziel steht die Entwicklung einer praxisorientierten Lösung für den sicheren Umgang mit produktionsnahen Daten in verteilten Systemen unter Anwendung der Blockchain-Technologie. Als Beispielszenario findet diese Technologie Anwendung bei der Überprüfung und Kalibrierung einer Werkzeugmaschine mit einem Double-Ball-Bar-Messsystem (DBBM). |
Objective
Motivation
The inspection and calibration of machine tools involves a great deal of effort. The machine operator incurs costs by commissioning the machine manufacturer or companies specialized in such measurements. In addition, the associated machine downtime of several hours to days results in expenses.
Objective
The following objectives are pursued in the project:
- Combination of product (measuring device) and service (machine calibration)
- Augmented reality (AR) based assistance system for user guidance during data acquisition
- Establishment of a scalable blockchain-based service platform to create legal certainty when dealing with data in machine and plant construction
- Complementing existing modeling procedures with machine learning algorithms to determine calibration parameters while taking environmental factors into account
Solution
In order to significantly reduce the costs of machine calibration, the machine operator must be enabled to perform the necessary measurements himself and to commission the evaluation of the determined data as a service. As a suitable measuring device for determining the machine parameters required for calibration, a DBBM is being further developed on the basis of an existing prototype for simple and robust use.
Figure 1: DBBM in use on a DMU 80 eVo
To enable the machine operator without additional qualification to use the measuring system correctly and to perform the measuring run according to the service provider's specifications, an AR assistance system is implemented for user support.
Figure 2 (left): Full 3D models machine METROM P800 and DBBM on calibration standard; Figure 3 (right): Reduced, AR-suitable 3D models.
A scalable digital service platform based on distributed ledger technology (DLT) will be used for the secure transmission of company-specific data and the legally compliant processing of the calibration services to be performed.
Figure 4: Overall concept
Results
The result is an easy-to-use, robust measurement tool for determining and improving the spatial motion accuracy of a machine tool. The measurements themselves can be performed by the machine operator without additional qualification through the use of an AR support system. The calibration service is handled via the digital service platform. The user commissions the service provider via the platform, performs the measurement run with the DBBM using the AR assistance system, and then transmits the measurement data. The service provider checks the data for plausibility and performs the calculations required for calibration. He then transmits the new calibration parameters to the user, who uses them to update his machine. This procedure greatly reduces the cost of calibration by eliminating travel expenses for the service provider and greatly reducing machine downtime.
Starting from a prototype developed in-house, the DBBM was further developed and the interfaces to the machine tool were defined. Based on the requirements of a measurement for a calibration, the need for support and the forms of interaction for the user were worked out and the corresponding procedure model was derived with the project partners. This information forms the basis for the AR assistance system to be developed.
Contact
© Crispin-Iven Mokry
Research associate
NameMr Dipl.-Ing. Frank Arnold
Process Informatics and Machine Data Analysis
Send encrypted mail via the SecureMail portal (for TUD external users only).
Chair of Machine Tools Development and Adaptive Controls
Visitor Address:
Kutzbach-Bau, Room E5 Helmholtzstraße 7a
01069 Dresden