Time-series and sensor AI.

Technologies

• Anomaly detection on sensor streams Detecting faults, events and drift in continuous sensor data at the edge.
• Forecasting and nowcasting Short-range forecasts of weather, water, load and traffic from local and regional observations.
• Predictive maintenance Estimating remaining useful life of equipment from vibration, temperature and electrical signatures.
• Sensor quality control and drift correction Automated checks and corrections that keep observations fit for use.
• Seismic and acoustic event detection Picking and classifying events in seismic and acoustic records.
• Hydrological time series Learned rainfall-runoff and water-level models using standard water data exchange.
• Network telemetry analytics KPI anomaly detection and prediction for RAN, core and transport.
• Virtual and soft sensors Estimating quantities that are not measured directly from other sensors and models, labelled as estimates.