How to Run AI Agents at the Industrial Edge
Jason Farmer, Product Manager on Microsoft’s digital operations team, walks through the infrastructure behind physical AI and how it supports virtual coworkers for industrial teams.
Connecting industrial equipment to AI agents
Physical AI connects real equipment to AI agents through a hybrid cloud-and-edge architecture. The connection path depends on where devices operate:
- Cloud-connected devices connect directly to Azure IoT Hub.
- On-site devices connect through Azure IoT Operations, running on Kubernetes at the edge.
- Foundry Local runs AI workloads on premises.
- Data from both device paths flows into Microsoft Fabric.
- Azure Resource Manager provides management across the architecture.
This arrangement supports both cloud-connected equipment and equipment that needs local infrastructure.
Virtual coworkers at the edge and in the enterprise
Farmer explains the architecture through two types of virtual coworkers:
- An operational technology (OT) virtual coworker runs on site and continues working even when the cloud connection drops. Local execution supports continuity for on-site operations.
- An enterprise virtual coworker runs in Microsoft Foundry or Fabric and reaches frontline and information workers through Teams and Dynamics.
The distinction is where the AI operates and how workers access it: the OT coworker supports on-site work with offline continuity, while the enterprise coworker connects to the applications used across the business.
Video chapters
- 0:00 — What physical AI needs at the edge
- 0:40 — How cloud and on-site devices connect to Azure
- 1:25 — How physical AI data flows into Microsoft Fabric
- 1:55 — What virtual coworkers do for frontline teams
- 2:20 — How an OT virtual coworker keeps working offline
- 3:05 — Where to see physical AI in action