Nexma

Monitor an area of interest

Imagery to alerts

In this tutorial you will set up continuous monitoring for an area of interest — from drawing the AOI to tasking fresh imagery, detecting change, and getting an alert when something moves. You will learn how to create a project, define an AOI, task imagery with Nexma SatelliteEngine, run detection and change detection with Nexma InsightEngine, surface results as live data on the Globe, and set an alert automation. The only prerequisite is an account in an organization; see Quickstart if you need one.

What you'll build

A standing monitor over one AOI: a fresh imagery task, an object-detection pass and a change-detection pass against a baseline, detection results rendered as live spatial data on the Globe, and an automation that re-images on a schedule and alerts you when meaningful change appears. You will end with a running monitor you can leave in place.

1. Create the project

From Projects → New, name the project (for example, "Port Watch") and bind its world model:

  1. Pick an imagery monitoring ontology — AOI, ImageryTask, Detection, ChangeEvent, and related entities.
  2. Add an imagery skill that targets it, carrying the detection model bindings and change-detection rules.
  3. Click create. The project opens with its own Globe view and Jax session.

If you would rather describe the work — "watch this area for new construction" — Jax proposes a matching ontology and skill. See Projects and Skills overview.

2. Define the AOI

Open the Globe and search to your area. Draw a polygon around exactly what you want to watch — a port, a corridor, a parcel block. The AOI is saved into the project as a real entity that every later step references.

Keep the AOI tight. A smaller footprint means cheaper imagery tasks, faster detection, and fewer false positives from areas you do not care about. You can edit the polygon later, or maintain several AOIs in one project.

3. Task fresh imagery

Open the Jax panel and ask for new imagery over the AOI:

Task fresh imagery over the AOI. Prefer optical at sub-meter resolution with less than 10 percent cloud cover. If clouds block it, fall back to SAR.

Jax uses Nexma SatelliteEngine to find and task the best available capture, choosing optical for clarity or SAR when cloud cover or night would defeat optical. When the capture arrives it renders on the Globe as a layer over the AOI.

You also want a baseline to compare against:

Pull the most recent prior capture over this AOI as the change-detection baseline.

4. Detect and detect change

With imagery in hand, run analysis through Nexma InsightEngine. Start with object detection:

Detect vessels, vehicles, and structures in the latest capture over the AOI. Surface each as a detection with a confidence score and location.

Then run change detection against the baseline:

Compare the latest capture to the baseline and flag meaningful change inside the AOI — new structures, removed structures, and large stockpile changes. Ignore seasonal and lighting differences.

InsightEngine returns detections and change events as structured records tied to coordinates inside the AOI. Push on them:

  • "How many vessels are in the AOI now versus the baseline?"
  • "Show me the largest new structure and its footprint."
  • "Filter to detections above 0.7 confidence."

See Capabilities: imagery.

5. Surface results as live data

Detections and change events render on the Globe as live spatial data — points and footprints over the AOI, styled by type and confidence. Click any detection to read its full record: model, confidence, capture time, location. Right-click to pin it into the Jax conversation as context for a follow-up question.

Because results are stored as project entities, they accumulate over time. Each new capture adds a layer of detections, so the Globe becomes a history of how the AOI has changed, not just a single snapshot.

6. Operate — set an alert automation

A monitor is only useful if it watches without you. Make the loop recurring:

Create an automation that re-images the AOI every 24 hours, runs detection and change detection against the previous capture, and sends an alert if any new structure over 100 square meters appears or vessel count changes by more than five.

The automation tasks imagery, runs InsightEngine, and notifies you only when its thresholds are crossed — so quiet days stay quiet and real change reaches you fast. Tune the thresholds anytime by asking Jax to adjust them. See live data.

When an alert fires, reopen the project: the triggering capture, detections, and change events are already on the Globe, ready to inspect.

Recap

You created a monitoring project, drew an AOI, tasked fresh imagery with SatelliteEngine choosing optical or SAR by conditions, ran object detection and change detection against a baseline with InsightEngine, surfaced the results as live data on the Globe, and set an automation that re-images on a schedule and alerts on meaningful change. Every step was a real, reversible change grounded in the project's world model.

Where to go next

Monitor an area of interest
Edit on GitHubSend feedback