Retailers across Britain are turning to Artificial Intelligence (AI) to help transform their energy efficiency, improve their carbon footprint and reduce their operating costs.
Central England Co-operative, which runs stores across the Midlands and East Anglia area, is using AI to go beyond consumption analytics towards a system that progressively ‘learns’ what optimal energy performance looks like in each of its retail outlets.
Over a three-month trial led by AMRdna, an Energy Assets service, the Central England Co-operativeproject generated a 206% return on investment by identifying and eradicating energy waste and implementing evidence-based efficiency strategies.
The project applied AI algorithms developed by kWIQly to analyse huge volumes of historic consumption data from across the Central England Co-operative portfolio to assimilate daily consumption data into an optimal performance model for each store. This makes unusual energy usage spikes much easier to spot…and address through remedial action.
As a result of this work, AMRdna identified numerous opportunities for Central England Co-operative stores to reduce wasted energy, including flagging two events that had resulted in unnecessary energy consumption valued at over £3,500. Additional, suggested actions by AMRdna led to electricity savings worth over £4,000 through improved efficiency.
The core value of this AI technology is its ability to assimilate meter data in a way that enables it to ‘learn’ what optimal building energy efficiency should look like. Creating this profile enables energy events outside this profile to be quickly identified and remedied.
“Energy waste can result from things as simple as leaving the lights on or heating overnight,” says George Catto, Client Services Director at AMRdna. “Our algorithm transforms consumption data into benchmark energy performance within individual buildings and across portfolios – and when exceptions occur, the system flags a deviation.”
Energy managers have historically used automatic monitoring and targeting systems to benchmark performance, but this does not offer true insight into optimal energy performance, only variation over manually set metrics.
The Central England Co-operative project involved the application of a simple workflow tool that guided energy managers through a process based on the identification, investigation and resolution of energy exceptions. Flagged problems can sometimes be as simple as lights being left on overnight, but in a couple of instances, the system identified that the Co-op was still being billed for energy at sites it no longer operated.
The AI also adapts to changes in store operation, for example where new freezer equipment is installed or if the site moves to 24-hour operation, so that only events outside the new ‘normal’ are flagged.
In total, the three-month pilot project resulted in 40 recommendations for improvements, 14 of which were quickly implemented, with more identified as longer-term fixes.
Luke Olly, Energy and Environment Lead at Central England Co-operative, commented: “The project led to an improved understanding of the energy performance of our sites, enabled us to identify issues more quickly and to set a consumption variation threshold for investigation. It also contributed to improved communication between head office and local stores and increased awareness of the direct effect that behaviours can have on energy efficiency.
“Indeed, the response was very positive from stores, who were able to investigate and address any issues more quickly than previously and were appreciative of the insight that enabled them to optimise energy performance.”
Central England Co-operative is continuing its relationship with AMRdna and will be adopting the technology within its standardised energy management processes, while also looking at the value the data-driven AI can provide to more strategic projects across the organisation’s store portfolio.
To find out more about AMRdna and how it can provide cost saving opportunities for you, click here.