HonAI: AI-enabled predictive stabilization and optimization of honey processing

Challenges addressed 

With increasing manufacturing volumes and several lines in operation, we wanted optimize energy consumption and reduce energy expenses. Due to unequal capacity usage we wanted to regulate uncontrolled energy peaks, which would allow us to operate at lower connection power. Also in terms of quality, Medex’s Honey Syrups, our flagship products, needed more stringent in line and real time control, since we had several, unpredictable problems with lid closure, resulting in quality non-compliant product and lost batches or time consuming rework. Utilizing the modular software platform, enhanced with AI models and algorithms, we predicted the quality of Honey Syrup bottling, improving product quality, process stability, and production efficiency while reducing energy consumption.

Solution

We enabled insights into energy streams and forming the basis for optimization to achieve “green” production by introducing energy efficiency model, which forecasts production energy consumption, identifying anomalies for quicker responses. We introduced two AI models—one for energy efficiency and another for operational efficiency. We integrated quality monitoring with the bottling machine which allows us to monitor the bottle lid closing momentum and predict non-compliance with AI. We also digitally organized and monitored maintenance activities to minimize unplanned downtimes and reduce breakdowns

Impact

• With the help of the Dynamic pricing (Peak shaving) module, we managed to reduce costs, induced by electricity power peaks by approximitely 10%, but further reductions are still possible and in the works with module optimization
• In addition to costs, the lower power consumption contributes to a lower power grid strain and in turn can have ecological effect on a larger scale
• Heat consumption reduction of 9 % was also achieved by reworking the heating device schedules and lowering the heating supply temperatures into the premises while not lowering the actual ambient room temperatures. 

The project gave an opportunity to collaborate with Opticum d.o.o., which was a great partner and everything ran smoothly and a lot of new knowledge was shared, giving as insight in further possibilities to enhance our operations with digitalization and AI. We look forward to cooperate on further developments.

During (and as a consequence of) the project, user awarness about energy saving was also raised. By simple behavioral changes (shutting off the lights, computers, A/C devices, electricity consumption was lowered by 3 %.

There are several further plans to dig deeper into digital solutions. We have many processes still to cover, from production planing using digital twins, to sales planning and AI enabled forecasting, to prevent out of stock moments or overstocking. New developments are planned also in managing advanced quality control by upgrading existing quality control module. We also plan digitalize manufacturing flow and in-process control to enable better monitoring and faster response to non-compliant actions during operations. 

Testimonial