World's first automatic acceptance test of a production machine for the beverage industry at an OEM in the collaboration between Coca-Cola HBC and the IIoT company Quantis
In mid-January 2022, the world's first automatic data validation of a production machine by means of a Factory Acceptance Test took place at an OEM at the request of Coca-Cola HBC.
Coca-Cola HBC was founded 70 years ago. Today, it is one of the world's largest strategic bottling partners of the Coca-Cola Company with operations in 29 countries on three continents.
Coca-Cola HBC has set itself the goal of rigorously implementing its digitalization strategy in the coming years. The company is actively investing in new technologies to simplify processes, always ensuring that these solutions support its Growth Strategy and customer centric approach.
For plant operators such as Coca-Cola HBC, it is always a challenge to obtain valid data on the utilization and efficiency of their plants, because in most cases the sensor machine data is not calibrated or insufficiently calibrated, and thus data is transmitted incorrectly or not at all. Therefore, Coca-Cola HBC decided to perform an automated Factory Acceptance Test in October 2021 before the delivery of newly ordered machines.
The Factory Acceptance Test
Quantis has developed PureData, a globally unique tool to validate the data specified by Coca-Cola HBC. For this purpose, the specified data points were stored in PureData with the associated test criteria, rules and AI methods.
An automated Factory Acceptance Test usually runs over a time range of 8 hours. The machine under test is connected to the PureData cloud application via the Internet. The baseline data of the machines was continuously transferred to the Quantis platform "Pocket Factory" in real time during the 8-hour test period.
The incoming data streams from the machine were checked for validity using the PureData application. For this purpose, in addition to basic rules, more complex dependencies were mapped in the system for these KPIs’. These were then checked under the rules and AI methods created. At the end of the test, a quality score was produced for each data point (KPIs’), which reflected the quality of the data point to be validated.
In this way, it was possible to identify data points that were not provided by-machine. Parallel testing of discrete data, counters, process variables, alarms and fault messages was performed. After completion of the Factory Acceptance Test, some data points were found to have non-valid values. In particular, the operating state required for calculating OEE was not available in sufficient quality. This was detected during the Factory Acceptance Test and helped the OEM to improve before delivery of the machine. As a result, Coca-Cola HBC can quickly start collecting validated data as soon as its plant is commissioned.
After evaluation of the data, all findings were logged and provided to Coca-Cola HBC in the form of a detailed results report. The OEM then made the improvements before the machine was delivered.
Even before the machine is implemented, high data quality can be ensured with the help of the automated Factory Acceptance Test. No rework is required at the plant operator's site. This saves both the plant operators and the OEMs high follow-up costs. With this, our customer Coca-Cola HBC has gained a certainty on the completeness and validity of his machine data.
Quantis provides plant operators with PureData, a unique tool that helps to check the data quality of machines even before the production plant is commissioned, thus avoiding expensive rework.