Energy / Oil and gas

Williams: Helping frontline workers find procedures and receive early equipment-anomaly alerts

Company
Williams
Country
United States
Adoption stage
In operation
Source published
Date basis
The date the source was published. It can differ from the date adoption started.
How the source was checked
Read the full source text

The work problem

Microsoft's customer story says frontline workers at natural gas company Williams could spend hours or days in the office finding permits, maintenance manuals and forms. Turbine efficiency was assessed by visiting equipment, reading parameters and making an experienced judgment. The company needed more time for inspections and repairs and earlier warning of equipment problems.

Technology and data

Williams provided Microsoft 365 Copilot to frontline staff and let employees create agents for their own workflows. An orchestrator connects those capabilities for others to use. Operations technician Andrew Allen built an anomaly detector using Copilot Cowork to retrieve equipment reports from an Outlook distribution list and compare them with baseline operating statistics. It alerts on changes such as rising thermocouple temperatures, falling oil pressure and increasing vibration; the technician investigates and acts. Copilot also supports plans, procedure interpretation and safety reports.

Results

Microsoft's story reports that AI adoption at Williams increased from single digits to over 90%. The company described less document searching and manual work and earlier identification of equipment issues before they became costly or dangerous. The adoption figure covers the organization, not the detector's accuracy or reduction in equipment failures.

Limits and open questions

False-alert rates, prevented failures and hours saved were not quantified. The story does not say AI directly controls equipment.

Sources

Compiled from public sources. These are not results from ATF Works customers.

Read original (opens in a new tab)