A suite of statistical tools have been developed by researchers from the University of South Wales for a spin-out company, RUMM Ltd (Remote Utility Monitoring and Management), to analyse energy usage for businesses in the manufacturing and utility sectors and automatically alert unusual activities, e.g. excessive consumption. The statistical tools comprised a key component of RUMM Ltd’s intellectual property and was central to the company’s purchase by RWE npower Ltd.
Statisticians at the University of South Wales have a long history of applying time series forecasting to energy usage, starting 20 years ago through projects with South Wales Electricity Board (SWALEC). As a result of this expertise, our researchers were approached by a University of South Wales spin-out company, RUMM Ltd, to further develop their analytical techniques in detecting irregular behaviour on energy usage.
RUMM Ltd was an energy management company whose clients were commercial businesses in the manufacturing and utility sectors seeking to reduce their energy bills and carbon emissions. Of particular interest for RUMM Ltd was the detection of excessive and irregular energy usage; for example, as caused by machinery being left on at unnecessary times.
Our statisticians, with assistance from doctoral research students, developed a suite of statistical tools utilising a Hierarchal Profiling Approach (HPA) that formed a key component of RUMM Ltd’s intellectual property. Within a few years, RUMM Ltd were identified as a thought leader in energy management and were courted by major energy companies. In 2015 RUMM Ltd was sold to RWE npower Ltd who subsequently incorporated RUMM’s technologies to benefit its customers.
At the point of sale, RUMM Ltd had saved its customers £43 million in energy bills, reduced energy consumption by 600 million kWh with a concomitant reduction of carbon emissions by almost 300 000 tonnes.