Facilities Management IoT platform - Predict the efficiency of equipments and energy consumption
Xen.AI team has been working with a leading facilities management company to add few machine learning features to their IoT based facilities management platform to manage the buildings and properties. Our goal was to predict the effectiveness of few equipments running in the client buildings, namely for Chillers, Heat exchangers, and Heat Returning wheels (HRW).
For each building, we have analysed anomalies in the energy consumption. These anomalies (observed mainly as a function of time of day and day of week) have been reported to the client. All the forecasts for more than 40 equipments have been generated daily using two different models based on the time series data. Final prediction is generated using a weighted average of the two models, where the weights are caused by model accuracies. We run daily control scripts that provide us a feedback on each model performance. We also set a number of logs and messages that control prediction status and generate alerts in case of any significant problems. The Xen team participated in the project is spread across the globe. However, still demonstrated high efficiency and timely delivery.
Project summary: AI and Machine Learning solutions for Facilities Management IoT platform to predict the efficiency of equipments, forecast the energy consumption and predictive maintenance of equipments.