Resource efficiency and maximum production uptime are top priorities for food manufacturers. Led by Hungarian REACH Solutions, the FEB project aims to support both these goals by validating a smart data management system that is designed to log all events, including errors, warnings and messages, sensor states and drive parameters via logical controllers. Using this data, the system can identify production anomalies, enable root-cause analysis and optimise component usage.

REACH Solutions is collaborating with GoodMills Ltd to overcome issues with downtime and quality. The goal is to increase the mill’s current OEE from 75% to 80% through more organised maintenance and downtime prediction.

If successful, the data management system will increase effective production time from 6.5 hours to 7 hours per shift, reduce energy consumption by 4% and cut waste due to improper packaging by 8%.

Robot-X Hungary and Óbuda University are also supporting this project.

Final report summary
The FEB project has achieved its goal to develop and validate an easily customisable platform with integrated functionalities.

After an initial survey of the selected use case at GoodMills Hungary’s packaging plant, the first tasks were to localise controllers on the production line, create a data collection network and build the application environment. Device-level data collection is performed by a future-proof telemetry component. Data is then stored in an effective, time series database.

The database enables data checking and monitoring functions that notify maintenance/management teams when there are production stops or if packages are under or over-filled.

A machine learning environment was developed with a focus on predicting downtimes. This has two aspects: to create a predictive maintenance tool and to identify the most important errors and their properties.

The final task was to create a tailored user interface that provides easy access to stored production information on computers and smart phones. User interface functionalities include real-time OEE monitoring, automated production planning, downtime registration, downtime and OEE reporting and specific alert notifications. When a breakdown occurs, the interface also highlights where the problem has occurred on the line.

Following implementation of the system, OEE on the selected GoodMills packaging line increased from an average 62% to 80%. Real-time monitoring of OEE has replaced the previous manual method of OEE calculation. Energy management has improved and flour loss reduced due to the sharp focus on overfilled bags.

GoodMills has now placed an order to install the IoT system on another packaging line at its plant and plans to introduce it to other locations. Further functional enhancements of the system are in progress.