Platform Integrations Case Studies Pricing Blog Contact Sign In
Platform

Feeder-level intelligence. Grid-wide control.

Ampgrove combines a rolling load forecast engine, a threshold-based dispatch layer, and a configurable alert system into one integrated platform built for the way distribution operations teams actually work.

Load Forecasting

Multi-horizon forecasting at the feeder, not the substation.

Most EMS load forecasts operate at the substation level and look 24 hours ahead. That resolution is too coarse to catch the overload building on a single feeder over the next four hours.

Ampgrove produces rolling forecasts at 15-minute, 1-hour, 4-hour, and 24-hour horizons for every feeder in your network. Models train on your historical interval data and continuously retrain as operating conditions change.

  • 30-minute resolution forecast updates
  • Weather-adjusted using local NWP inputs
  • Model retraining on rolling 36-month history
  • Uncertainty bands per forecast horizon
  • Accuracy reporting by feeder and season
Abstract network topology visualization showing interconnected nodes representing feeder connections
Dispatch Recommendations

Recommendations that surface when operators need them, not after.

When a feeder's rolling forecast crosses a configured load threshold, Ampgrove generates a specific dispatch recommendation: which resources to activate, how much to shift, and what the expected load reduction is at the feeder level.

Operators review and approve. Ampgrove never executes a dispatch action automatically. The audit trail captures every recommendation, the operator's response, and the resulting load outcome.

  • Threshold-based triggers configurable per feeder
  • Operator override workflow built in by design
  • Full audit trail of recommendations and responses
  • Resource availability check before recommendation
  • Post-dispatch load outcome tracking
Abstract data flow visualization representing continuous dispatch signal streams
Alert Engine

Alerts tuned per feeder so each notification means something.

Alert fatigue is an operational risk. Ampgrove's alert logic is feeder-specific: thresholds, escalation paths, and notification channels are configured independently for each feeder in your network.

Configurable Thresholds

Set load warning and alert thresholds independently per feeder. Summer peak feeders carry different limits than residential distribution feeders. Ampgrove does not apply site-wide defaults.

Multi-Channel Notification

Route alerts to email, SMS, or paging systems depending on severity and feeder. Critical-path feeders can have dedicated escalation chains separate from standard monitoring alerts.

Transparent Alert Logic

Each alert includes the forecast confidence score, the sensor reading that triggered it, and the recent load trend. Operators can assess whether to act immediately or defer based on real context, not just a threshold breach number.

Data Pipeline

Your SCADA data arrives. Ampgrove normalizes, validates, and models it.

The data ingestion pipeline is where most utility AI projects stall. Ampgrove ships with pre-built connectors for common SCADA and EMS platforms and handles the normalization work that typically takes months of custom engineering.

Step 1 SCADA / EMS Read
Step 2 Interval Normalization
Step 3 Weather Data Merge
Step 4 Model Inference
Step 5 Forecast Output

Get a pilot configuration for your grid.

We scope your feeder topology, configure thresholds, and run a 30-day pilot on a subset of your feeders. No upfront commitment.