AI Predictive Analytics
ML models that predict equipment failures before they happen — with >83% accuracy
Our predictive AI engine analyzes vibration, acceleration, and temperature data from connected sensors to detect anomalies and predict failures — weeks before they occur. Developed in collaboration with A*STAR and SIMTech, these models use unsupervised learning for anomaly detection and semi-supervised soft clustering for multi-class fault prediction. Currently powering predictions across our AFOS Fusion products, with less than 25% customization effort required for new equipment types.
Enterprise AI Knowledge Base
Ask any question about your facilities — in plain English — and get an instant, accurate answer
Athena is an AI assistant purpose-built for facility operations. Powered by Retrieval-Augmented Generation (RAG), it indexes your equipment manuals, maintenance procedures, compliance documents, and historical fault data — then answers questions in natural language. Embed it in any dashboard, mobile app, or deploy it standalone.
- RAG-powered enterprise knowledge retrieval
- Natural language query: "What's the procedure for replacing the drive belt on escalator model X?"
- Indexes: equipment manuals, SOPs, compliance docs, historical fault records
- Three deployment modes: embedded in IRIS, embedded in any AFOS product, standalone
- Reduces dependency on senior technician expertise
- Training tool for junior staff
Digital Twin Platform
Web-based 3D visualization and simulation — see your building live, simulate its future
Softgrid's Digital Twin platform turns your physical assets into interactive 3D models with live sensor data overlay. Powered by Unreal Engine and delivered via Pixel Streaming, it runs in any web browser — no installation, no special hardware. Use it for real-time monitoring visualization, "what-if" scenario simulation, or structural health monitoring with real-time finite element analysis.
- Unreal Engine + Pixel Streaming (cloud GPU rendered, browser-based)
- 300,000+ data updates per second
- Real-time Finite Element Analysis (FEA) visualization
- Live sensor data overlay on 3D building/asset model
- "What-if" scenario simulation for maintenance planning
- Structural Health Monitoring for high-value assets
- No client-side installation required
