Use Case: Predictive Hospital Campus Resilience During Severe Weather

DEMO

Predictive Hospital Campus Resilience During Severe Weather

Predictive AI protects a hospital microgrid from severe storms, pre-charging batteries and staging operator-approved load shedding so critical care power never drops.

Building Blocks used:

ai.agents

automation.optimize

Skills used

Download skills for IBM Bob →

Products Used

1 Review use case

A severe storm approaching a hospital campus puts critical operations at risk. When solar output collapses, a grid outage could leaves the battery system unable to carry the required load to support the acute-care departments, relying on Hover Wind. A dashboard that only alerts operators after the fact is not enough. Facilities teams need to know hours ahead what the campus energy usage will look like and what to do before the disruption arrives.

Hover Energy's Microgrid Management System (MMS) combines IBM Environmental Intelligence Suite with IBM Maximo's asset telemetry and IBM Granite time-series to predict to forecast weather, generation, demand and capacity gaps across the campus. IBM watsonx Orchestrate's agents support interactive exploration of system decisions and recommendations.

This demo simulates an approaching storm highlighting solar, wind and battery metrics. When a grid outage occurs, battery backup and wind generation prevents clinical impact while a Maximo work order addresses a degraded inverter. Personas include the facilities/energy manager who approves the plan and a clinical operations lead who needs assurance that patient-care loads are never impacted.

Start Here

Watch this short overview

Preview the application

Explore key screens from the demo

2 Try the interactive application

Experience the interactive demo and explore its features. A demo script is provided to help when presenting the application to partners.

Try the demo

Have questions, want a walkthrough or found a bug?

Contact the demo owner for assistance.

Submit a request →

3 Explore how it was built

The demo layers six components. IBM Environmental Intelligence Suite supplies hyper-local storm and weather forecasts. Edge telemetry from PV, wind, and battery assets flows through IBM Maximo Application Suite, which tracks asset health and generates work orders. IBM Granite time-series models on watsonx.ai forecast generation, load, and capacity gap 96 minutes ahead. The Hover Energy MMS (FastAPI telemetry and fault-detection services over a SQLite dataset) applies deterministic rules: a locked clinical protection floor, battery reserve calculation, and nonclinical load-shed planning. IBM watsonx Orchestrate hosts the agent that explains recommendations and routes them to a human operator for approval. A Next.js/Carbon UI proxies all calls to the backend services. In this demo, weather, generation, and battery data are simulated to stand in for the live feeds.

Architecture

4 Download and modify with IBM Bob

Download the source code and specifications for this application to set up and run it locally. Once downloaded, open the code in IBM Bob to modify and adapt it to your own use case.

Source Code

Step 1 — Review the Campus Overview

Open the Overview tab.
Show the three-building campus (A: acute care, B: outpatient, C: support).
Point out the 1,080 kW protected floor for Building A. All buildings are GREEN.

Step 2 — Explore the Weather & Forecast

Open the Weather & Forecast tab.
Trace the 7-day forecast and the Storm Week Profile.
Note that the curves are simulated for the demo. In production they are driven by IBM Environmental Intelligence Suite and Granite forecasts. The grid outage is a what-if scenario, not a prediction.

Step 3 — Highlight Autonomous Battery Pre-Charging

Point out the purple dashed line on May 4.
Explain that the MMS autonomously pre-charges the battery to 85% when storm risk crosses its threshold, and the battery holds that level until the storm.

Step 4 — Run the Storm Demo

Return to the Overview tab and click Run Storm Demo.
Building A goes AMBER as the storm approaches and inverter MG-17 degrades 11%.

Step 5 — Review the Storm Impact

Show the Storm Impact view: solar decays to 0 while the wind turbine keeps generating.
The AI flags grid-loss risk: battery plus wind must carry the 1,080 kW floor.

Step 6 — Review Alerts & Actions

Open the Alerts & Actions tab (red badge).
Review the autonomous battery staging.
Review the human-in-the-loop plan to shed 215 kW of nonclinical load with 0 kW clinical impact.

Step 7 — Approve and Execute the Plan

Click Approve & Execute Plan. Demand steps down by 215 kW.
At 14:50 the grid fails and the campus islands at the 1,080 kW protected floor, with battery plus wind carrying it until the grid returns at 16:30.
The campus returns to GREEN with a Maximo work order synced.

Step 8 — Ask the Agent

Open the Agent tab.
Ask the watsonx Orchestrate agent why the load shed was recommended.

Step 9 — Close on the Architecture

Close on the Architecture tab to show how the pieces fit together.