Use Cases for Fast Machine-Learning Algorithm

Industry examples employing FADAR

Use Case #1: Rooftop Unit (RTU)

Equipment such as roof-top air conditioning units (RTU) involve refrigeration fluid circulators, fans, and compressors, so energy use patterns of these systems can be complicated. FADAR can analyze the data, and through rapid iterations can arrive at a normal-state profile. As equipment continues operating, FADAR can then identify when energy use is out of normal parameters, and can trigger alerts on equipment conditions. Predictive analytics can indicate when the unit will require service such as a filter change or a change to the defrost cycle, saving money on maintenance and repairs. In addition, it can determine deterioration of performance before a failure requires major investment in new equipment. By prolonging the life of equipment and alerting on conditions before failures happen, FADAR can save on energy costs, maintenance costs, and the major cost of a catastrophic failure.

Use Case #2: Refrigeration

Schedule defrost times based on patterns of use rather than set times.

FADAR can analyze patterns of defrost cycles and temperature cycles to determine ideal defrost time and frequency. Refrigeration coils build up frost, reducing their efficiency. Defrost cycles in commercial refrigeration are usually based on set defrost times. Depending on the volume of the space, its insulation properties, and how often the door is opened, timed cycles can lead to more defrosts than are needed, using excessive energy. FADAR’s analytics can modify timed schedules dynamically to match current usage patterns.

Detect coolant leakage without special sensors

According to one expert*, the most common cause of coolant loss in commercial systems is inability to detect slow refrigeration leaks without expensive and often limited refrigerant sensors. FADAR can detect coolant leaks for any coolant without the need for coolant sensors, by analyzing energy use and temperatures. Coolant loss is costly in a number of ways, from reduced efficiency of the refrigeration unit, to increased energy required to maintain safe temperatures for food, to loss of refrigerated products. In addition, coolants containing hydrofluorocarbons (HFCs) and chlorofluorocarbons (CFCs) are potent greenhouse gases and CFCs can damage the ozone layer.

* Artificial Intelligence in Energy and Renewable Energy Systems, by Soteris Kalogiru, published by Nova Science Publishers, Inc., 2007

Use Case #3: Solar Equipment

Predict failure points in solar panels and water heaters

FADAR can detect and analyze decline in solar output in solar panels and use the data to project a future fail point with its predictive analytics. This can allow operators to schedule panel replacement by optimizing the timing to get the panel’s full life.

The same process can provide a lifetime estimate for solar water heaters, maximizing their efficiency during their lifetime. Based on declining performance, FADAR can project future failure, to save on maintenance and plan repair and replacement.

About the Inventor

Curtis Meadow, Chief Technology Officer for PowerWise, developed the algorithm. "We have always managed and analyzed huge amounts of data at PowerWise," says Meadow. "I created FADAR to help our customers with predictive analytics for their important business equipment and decisions."

Meadow received his Bachelor's and Master's Degrees in Computer Science from the University of Maine, initiating a career in database development, software development, natural-language processing, university teaching, and novel algorithm design.

PowerWise develops and sells energy and building management systems to help people reduce costs, better manage their facilities, and achieve energy and operational efficiency. The company is based in Maine. For more information, visit www.powerwisesystems.com or call +1-207-370-6517.