AI Field Service Knowledge Search Agent for Process Control Equipment

How Cassidy automates using AI

Step 1: Trigger on work order or alarm event

The Workflow activates when a new work order arrives from SAP PM, Maximo, or another CMMS/EAM system—or when a DCS/PLC alarm, predictive maintenance alert, or operator callout initiates a service request.

Step 2: Assemble asset context from the Knowledge Base

Cassidy pulls the relevant control narrative, P&IDs, loop sheets, wiring diagrams, device data sheets, and prior service notes linked to the asset tag using ISA-5.1 naming conventions, building a complete context package.

Step 3: Surface safety requirements and pre-job checks

The Workflow prompts PTW, LOTO, and JHA requirements aligned to site policy, including SIF bypass approvals and MOC checks for safety instrumented systems per IEC 61511.

Step 4: Deliver guided troubleshooting with offline support

Cassidy presents symptom-driven fault trees and OEM diagnostic flows, integrating HART/FF/PA device diagnostics and historian trend data—all cached for full offline operation in air-gapped environments.

Step 5: Capture calibration and proof test data

Structured forms collect as-found/as-left measurements, tolerances, and test results with digital signatures, automatically generating calibration certificates and SIS proof test reports.

Step 6: Sync closure artifacts and update knowledge

When connectivity returns, Cassidy syncs completed work—steps performed, measurements, device diagnostics, photos, and parts used—to the CMMS/EAM, closes the work order, and proposes new knowledge articles tagged for future retrieval.