Rudolfovo, together with Robotina d.o.o. developed the award-winning xEMS innovation

Photo: Media GZSS

Rudolfovo – Science and Technology Centre Novo mesto, together with Robotina d.o.o., participated in the development of the innovation Intelligent System for Optimising Energy Management in Buildings xEMS, which received the Gold Award for innovation in the Coastal-Karst region. The award was presented at a joint event organised by the Chamber of Commerce and Industry of Slovenia – Regional Chamber Postojna and the Primorska Chamber of Commerce, where the best innovations from the Primorska-Notranjska and Coastal-Karst regions were recognised. As one of the three best innovations from the Coastal-Karst region, xEMS also qualified for the national level, where it will compete for the national awards of the Chamber of Commerce and Industry of Slovenia.

Artificial intelligence for forecasting energy consumption in buildings

xEMS is an advanced building energy management system that uses artificial intelligence to provide accurate forecasts of energy consumption and solar energy production. It takes into account energy flows, battery status, building loads, weather data and electricity prices.

The system enables the optimisation of battery charging and discharging at 15-minute intervals. In this way, it helps reduce energy costs, increases the use of local renewable sources and supports more efficient building management.

The solution is intended for business, public, educational, hotel and other buildings, where better energy consumption forecasts can significantly contribute to lower costs and more sustainable operation.

The challenge: energy consumption is no longer only about quantity, but also about timing

Electricity demand in buildings is constantly increasing. This is driven by numerous consumers, including heating, cooling and ventilation systems, charging stations, digital equipment, building automation devices and other electrical systems.

At the same time, electricity is becoming more expensive, while tariffs are increasingly dependent on the time of consumption. For users, this means that it is no longer important only how much energy they consume, but also when and how they consume it.

Robotina therefore wanted to upgrade its energy management platform with energy consumption forecasts for one to three days in advance, at 15-minute intervals. Such forecasting requires advanced artificial intelligence models, high-quality data and substantial computing capacity. On standard equipment, training a demanding model could take several weeks, which limited the development and wider use of the service.

The solution: connecting data, artificial intelligence and high-performance computing

The partners developed a process that connects Robotina’s energy management platform with advanced artificial intelligence models and high-performance computing.

Data from smart meters was combined with weather data, electricity prices and other factors affecting energy consumption. Based on this, artificial intelligence models were developed to forecast future energy consumption in buildings.

By using powerful computing infrastructure, the time required for demanding model training and testing was reduced from approximately one week to just a few hours. The project also resulted in a lighter version of the artificial intelligence model, suitable for deployment directly on devices in buildings.

The role of the partners

The innovation was developed in cooperation between Robotina d.o.o., Rudolfovo – Science and Technology Centre Novo mesto and the Faculty of Mechanical Engineering of the University of Ljubljana.

Robotina d.o.o. contributed its energy management platform, knowledge of the real needs of building users and experience in developing solutions for smart energy system management.

Rudolfovo contributed expertise in artificial intelligence, data processing and the transfer of advanced digital solutions into practice. It participated in data preparation, the development of forecasting models and the validation of their applicability to Robotina’s real business needs.

The Faculty of Mechanical Engineering of the University of Ljubljana contributed expert knowledge in energy systems, the operation of buildings and the technical factors that influence energy consumption.

More reliable forecasts and concrete benefits for users

The innovation of the solution lies in enabling buildings to look into the future. Facility managers can see in advance when consumption will be higher and adjust the operation of devices, energy storage systems and other systems in good time.

This is particularly important in the context of dynamic electricity prices, solar power plants and battery storage systems. During the project, the forecasting error was reduced from 47.88% to 10.49%, resulting in significantly more reliable energy consumption forecasts.

For users, this means better energy planning, the possibility of lower costs, more efficient use of renewable energy sources and more sustainable building operation.

An innovation with development and market potential

For the development of the innovation, the consortium secured a European project. The project ranked among the top 10% of applications selected for funding, confirming its quality, innovativeness and market potential.

The Gold Award and the qualification for the national level confirm the importance of cooperation between business, research organisations and universities in developing solutions that address the key challenges of our time.