Artificial intelligence (AI) could be used by companies to help reach the voluntary 15% gas reduction target placed on EU Member States by the EU Council in July 2022, according to Arloid Automation, of London.

The company advocates the use of AI coupled with any building management system to control and make significant savings of up to 30% on energy and coolant use.

According to the Council of the EU’s press release (26 July 2022), ‘Member States agreed to reduce their gas demand by 15% compared to their average consumption in the past five years, between 1 August 2022 and 31 March 2023, with measures of their own choice’. The importance of reducing energy demand cannot be overemphasised largely due to supplies from Russia being significantly reduced, erratic and they may be stopped altogether.

Most energy saving measures have a significant capital outlay but Arloid says that its AI system can be implemented without cost until savings are made. The company claims it is one of the cheapest and easiest ways for EU states to significantly reduce energy consumption with no upfront costs. Once savings are established, after around 30 days, a percentage fee is charged.

The reduction in energy use is easily achieved through a series of steps. Firstly, the system produces a virtual building identical to the real one with the same construction materials, location, climate and personnel attributes. From this virtual building AI makes a series of simulations based on live data from the real building. This process takes about a month.

The AI intuitively adjusts the heating, ventilation and cooling settings resulting in real savings. In the real world building services engineers or property managers have previously adjusted these settings but this can take some time. The company says that by using its system these adjustments take place automatically to save money and make the building more comfortable.

Arloid’s AI uses Deep Reinforcement Learning to automatically manage the operation of HVAC (Heating, Ventilation and Air Conditioning) systems in a wide range of buildings via a secure Virtual Private Network (VPN). It then makes decisions based on reinforcement behaviour and real-time data to provide faster optimisation, better HVAC performance and reduced energy usage.