Accelerator Hub
AI / ML Models
Reduce costs, improve customer experience, and create top-line value using AI/ML powered insights.
Reduce bottom line costs, improve customer experience, and create top line value using AI/ML powered insights. Explore pre-built models engineered for IoT fleet and service use cases.
Available Models
Service Action Prediction
Predict the number of service actions for the next twelve months for a population of devices.
Supplies Loyalty Model
Utilises field data to accurately classify consumables as genuine or non-genuine supplies.
Exception Based Rules
Trigger alerts when specific device error patterns occur, enabling proactive responses to device failures.
Model Monitoring
A robust set of features focusing on monitoring data quality, data drift, and prediction drift across deployed models.
Event Anomaly Detector
Identifies anomalous events using period-aggregated data from a population of devices.
Autoencoder Model
A type of artificial neural network used to learn efficient codings of unlabelled data (unsupervised learning).
How Models Fit the Journey
AI/ML models in the Accelerator Hub are designed to integrate with the Connected Products use case progression:
- Descriptive — understand what happened using dashboards and telemetry data.
- Proactive — detect issues as they emerge using exception-based rules.
- Predictive — anticipate failures before they occur using trained ML models.
- Prescriptive — automatically drive the right action using model outputs combined with the rules engine.