Georgia-Pacific’s Fletcher Facility Develops AI Tool To Close Gaps in Safety Process

FLETCHER, Okla., March 25, 2025 /3BL/ - Georgia-Pacific’s gypsum facility in Fletcher, Okla., is experimenting with artificial intelligence to help make sure people go home the same way they came to work.

Maintenance planner Jimmy Stewart is developing an AI-powered app to enhance employees’ connection to GP’s safe work permitting standard, Save My Life (SML). The app, called SMLGPT, is meant to be used alongside the SML thought process and the original SML app to help the facility identify hazards and protect employees and contractors.

“I built this in order to help train others, especially newer employees.” Stewart said. “If they’re working after hours or there aren’t any safety personnel around, SMLGPT can help them get the answers they need.”

The SML thought process is built on three key questions:

1. When this fails, what will kill me or change my life?

2. What controls are in place?

3. Is that enough?

The original SML app guides employees and contractors through that thought process with a digital work permit template. The goal of SML is to ensure that the necessary controls are in place before work starts, both to prevent failures and to mitigate the impact of failures when they occur.

However, what if people don’t have the experience needed to know all the potential hazards that need controls? What if it’s the first time the facility is performing a task, so even veteran employees aren’t familiar with potential hazards? What if people understand the hazards but don’t know what controls could prevent or mitigate failures?

“We always asked contractors, ‘What could kill you or change your life?’” Stewart said. “And a lot of them didn’t know how to answer that question.”

SMLGPT is being trained to help close those gaps.

Employees will be able to upload photos and videos of a work area to SMLGPT. Or they can simply describe it in a prompt, like, “What hazards and controls should I consider in an area that has MOPED interactions?”

Then SMLGPT will help identify common High Potential Conditions (HPCs) and High Potential Events (HPEs) based on the prompt and recommend controls to protect people.

Stewart said he usually has the original SML app open on a tablet as he walks through the permit process, with SMLGPT open on another device to help answer questions or identify hazards.

Safety and health manager Chris Brown helps lead the implementation of SML across GP facilities. He said hazard identification is one of the biggest challenges for facility employees as they try to use the SML thought process. That’s especially true for newer employees with limited manufacturing experience who may not know what risks to look for.

“If you can take some photos or a video of an area, and the AI tool can help you identify the critical hazards, that’s huge,” Brown said. “And we would absolutely want workers to verify what AI provides as guidance, but I think it does bridge that gap between hazard identification and the permit.”

The facility is also experimenting with using SMLGPT to generate quizzes to assess employees’ understanding of SML and to recommend controls based on GP safety alerts. Meanwhile, Brown said Fletcher’s work with SMLGPT has helped his team think through how AI could enhance future versions of the original SML app.

It’s a timely example of how Principle Based Management™ can help GP realize its commitment to safety as its top priority. Stewart’s willingness to experiment with emerging technology will improve his teammates’ lives and close gaps, all while helping ensure people go home safely.

View original content here.