Manufacturers don’t have the luxury of stopping production every time someone needs to learn something new. Frontline employees need to get productive quickly, work safely, and adapt as processes change. Yet the knowledge they rely on is often buried in SOPs, job aids, presentations, and systems that are difficult to navigate in the moment they need them. And sometimes, the written process doesn’t even match what’s happening on the shop floor.
For Tai Davidson at Marvin, that’s where the trust problem begins. As Tai Davidson explains,
“If you want people to actually use it, you need them to trust it.”
When a new operator opens an SOP and it doesn’t match the machine in front of them, employees quickly learn to look somewhere else for the answer.
In this executive roundtable, leaders from Marvin, GAF, and Aarti Industries get practical about how AI can help close that gap. The conversation moves beyond AI theory into real frontline applications, from finding the right job aid in seconds and breaking down language barriers to creating assessments, keeping SOPs current, and capturing the knowledge of experienced operators before it walks out the door.
For Dr. Reena Sanghvi, accessibility is only part of the equation. Learning also needs to stick:
“They should be able to practice it to the extent that it becomes a muscle memory.”
On the shop floor, knowing something isn’t enough. Employees need to be able to apply it safely when it counts.
Erik Peterson at GAF takes that one step further:
“They should be able to practice it to the extent that it becomes a muscle memory.”
It’s a crucial distinction for frontline learning. Completing a module, reading an SOP, or passing a quiz doesn’t necessarily prove someone can perform the task when safety, quality, and production are on the line.
That’s the bigger question running through the conversation: how can manufacturers use AI to get critical knowledge into people’s hands faster, without removing the practice, experience, and human judgment that turn that knowledge into real capability?
Key takeaways
- Start with trust, not AI.
AI won’t fix unreliable knowledge. Get the SOPs right, establish clear ownership, and give AI trusted content to work from.
- Get the right knowledge to people faster.
AI can make large libraries of SOPs and job aids easier to navigate using everyday language, while voice, mobile, and translation can make that knowledge more accessible on the frontline.
- Capture the knowledge that isn’t written down.
Use AI to help turn the experience of skilled operators into practical job aids and video SOPs, then validate that knowledge with operations, safety, and quality teams.
- Turn existing content into learning people can actually use.
Transform existing materials into microlearning, video, job aids, scenarios, translations, and assessments that fit around the realities of frontline work.
- Move beyond completion to demonstrated capability.
Completing a module or passing a quiz doesn’t prove someone can do the job. Combine digital learning with practice, observation, and skills verification to demonstrate real capability.
- Don’t let AI shortcut the learning itself.
Instant answers are useful, but expertise still requires repetition, coaching, judgment, and experience. Use AI to support frontline learning, not replace the process of becoming genuinely skilled.
Who should watch
Manufacturing executives, HR and L&D leaders, frontline operations leaders, safety and compliance professionals, and technology decision-makers looking for practical ways to use AI to make critical knowledge faster to find, easier to use, and more relevant to the reality of work on the shop floor.
