EnterPrise | Use of a cyber-physical production system for efficient tool management.

In the EnterPrise project, the PTW is developing a cross-company solution in the form of a cyber-physical production system (CPPS) for the company AWB (tool user) in order to optimize their problematic tool management. The CPPS is intended to introduce both a traceability system based on auto-identification (autoID) and real time location systems (RTLS) as well as a machine learning model for predicting wear and tear with automatic, demand-optimized ordering processes. The project is carried out in cooperation with the tool manufacturer KOPP (supplier) and the IT company UHP.

Duration: Completed | 2020 – 2022
Funded by: HA Hessen Agentur: LOEWE-Förderlinie 3

If you have any questions about this completed project please contact our institute management:

Motivation

The strategic importance of Industry 4.0 is increasing in the manufacturing sector. The central concept of Industry 4.0 is the improved collaboration of all production participants, i.e. plants, machines, products, resources and people. The existing tool management systems of KOPP and AWB, occur to have improvement potentials within the management of their tool cycle. The large number of product variants leads to high stocks of cutting tools and frequent tool changes. On the one hand, large tool inventories tie up capital and reduce liquidity; on the other hand, the multitude of different tools makes their localization and condition monitoring more difficult. As a result of the undocumented tools and their storage locations, the required tools are missing as well as long search times, which lead to longer machine downtimes and increase the throughput time in production unnecessarily. The lack of condition monitoring leads to unplanned tool failures during the machining process and thus to rejects or rework.

The need to develop the solution sought in “EnterPrise” is essentially motivated by the expected increase in the competitiveness of the two medium-sized user and manufacturer companies AWB and KOPP through the targeted expansion of existing systems with the aforementioned functionalities.

Objectives

The aim of the “EnterPrise” project is to develop a cross-company solution in the form of a cyber-physical production system (CPPS) in order to optimize problematic tool management at manufacturers and user companies (AWB and KOPP). The CPPS is intended to record tool and order data and exchange them between the user and the manufacturer. AWB should benefit from the transparent remaining service life of tools through the use of the CPPS, so that an order-oriented JIT procurement management of tools is also enabled and the capital commitment is reduced.

For KOPP, this results in greater planning accuracy with regard to tool orders and regrinding orders from AWB, as well as a marketable service for tool wear monitoring for other customers. UHP and TUDa act as enablers for CPPS development. For this purpose, UHP takes on the creation of the necessary interoperable IT system architecture, while TUDa contributes its expertise to implement tool localization and identification as well as wear prediction

Approaches

The CPPS has two main tasks. On the one hand, it should record status information in the form of wear data on tools in AWB's inventory during their entire tool life cycle. On the other hand, the storage locations of the cutting tools should be mapped in real time. Clear identification and localization is required to track the condition and location of the tools.

The introduction of a hybrid traceability system is planned for this purpose. Central components are the development of a marking strategy, the selection of a suitable technology (data matrix code, RFID, UWB, etc.) and the formulation of requirements for the knowledge database (WDB) used at KOPP for the provision of status data for each tool (e.g. unique numbering system) .

In order to be able to make a recommendation for the use of tools in every production order based on the recorded status data, TUDa intends to develop an ML model that enables statements to be made about the wear behaviour of tools during machining. The first step is to identify such data (machine data, sensor data, etc.) that allow a statement to be made about this. In order to be able to give a realistic forecast of the wear behaviour during a production order, the process data assessed as wear-relevant for each machining step are recorded and then linked with the real tool measurement in order to train the ML model with it. The data ultimately serve as the basis for continuously recording and predicting the remaining tool life and assessing whether a tool change needs to be made.

Acknowledgement

The research project is funded by the HessenAgentur. We thank for the opportunity to work on this project.

Funding source

Consortium partner