Intelligence at the edge | THR905

Sean Parham and Isabella Calandriello present a Fabrikam industrial scenario that connects device security, factory asset modeling, and local AI-driven control. The demonstrations follow operational data from edge devices into plant information models and Microsoft Fabric.

Securing truck firmware

The session opens with Fabrikam identifying vulnerable truck firmware and preparing an over-the-air update using Azure Device Registry.

A software update job is created and deployed to the Seattle truck group through Azure IoT Operations. The demonstration then shows the trucks successfully updated to firmware version 1.5, with the identified vulnerabilities resolved.

Connecting factories and discovering industrial assets

Sean introduces the Azure Digital Operations Experience Portal for connecting factories and onboarding industrial assets.

The workflow shows how Azure IoT Operations:

Sight Machine builds a unified plant information model that connects machines, processes, and live operational data. This provides the context needed to relate individual device signals to plant operations.

Adding safety policies to the plant model

Data scientists enrich the information model with safety door policies and publish it through Azure Device Registry.

The demonstration connects this modeling work to a specific operational scenario: using the state of a safety door to determine how a pallet mover should operate.

Using local AI to adjust pallet mover speeds

The visual inspection demonstration uses AI and Foundry Local to detect safety door states. Those detected states feed into control of pallet mover speeds.

Azure IoT Operations handles the associated workflow, including:

The scenario brings together visual inspection, local AI, and operational controls rather than treating inspection results as a separate reporting output.

Bringing cloud and edge data together with Microsoft Fabric

The final segment shows Microsoft Fabric unifying operational data, semantic models, and AI-driven industrial intelligence across cloud and edge.

Together, the demonstrations connect several parts of an industrial workflow: updating vulnerable devices, discovering factory assets, building contextual plant models, enriching those models with policies, and using local AI results to influence equipment behavior.

Session chapters

Resources

This Microsoft Ignite 2026 session is THR905, an advanced-level theater session in the Data track, presented in English (US).