NANTUM AI
ANOMALY DETECTION
User Generated Alerts
No one knows their building better than the engineers and operators who work in it every day. Nantum allows a user to create an alert against any data point within the building, selecting the trigger threshold, severity level, and type of notification they would like to receive.
AI/ML Anomaly Detection
Every day, Nantum analyzes historical data and combines it with predictive expected behavioral analysis to create a predictive model for all the data points within the building. We call this a Dynamic Boundary. When a data point moves outside its Dynamic Boundary, operators are alerted to a possible anomaly within their building.
Semi-Supervised Anomaly Detection
Point Anomaly
If a data point can be observed against other data points as an anomaly, it’s a point anomaly. This is the simplest type of anomaly.
Contextual Anomaly
If a data instance is anomalous in a specific context (but not otherwise), it’s a contextual anomaly. An example would be abnormal water use in the middle of the night.
Collective Anomaly
If a collection of related data instances is anomalous with respect to the entire data set, it’s a collective anomaly. An example would be a BMS malfunction with multiple sets of abnormal data.
Data Quality
Analyze data accuracy, completeness, consistency, validity, uniqueness, and timeliness.
Impossible Values
Identify missing data or impossible values from all meters, IoT devices, and data points for team investigation.
ML Data Correction
Recover from data outages by utilizing ML-driven data correction with over 90% accuracy.
Meter Data Anomaly
Real-time meter data outage and Customer Success team support to investigate the situation.
BMS Anomaly
Real-time BMS anomaly detection, such as electric demand and steam spikes, current, and power quality alerts.
Occupancy Anomaly
Detect unusual occupancy patterns in real-time, like overutilized spaces and unusual activity late at night.
Air Quality Anomaly
BMS and tenant comfort management through real-time indoor air quality alerts.
Water Anomaly
Real-time water management and leak detection through water meters, and IoT devices.