ForgeShift
← All insights

AI in Manufacturing

Week of 19 October 2026

Agents · Floor Impact · Frontier Deployments

6 stories · curated by ForgeShift Advisory

Deployment

Fastenal's sales force shrank by 400 in Q3 as it leans on AI

Industrial distributor Fastenal's sales force shrank by 400 in the third quarter. Executives say they need fewer people at customer sites and are counting on AI, including a recent small acquisition, to keep improving efficiency.

A reported headcount outcome with AI named as the lever; the productivity result itself is not given.

Frontier

DeepMind's Gemini 4 Argon teaches Apptronik's Apollo humanoid from video

As part of launching Gemini 4 Argon, Google DeepMind demonstrated the model learning a physical task by watching video, then transferring the behavior to Apptronik's Apollo humanoid robot.

A launch demonstration, not a plant result; watch for deployments before it affects your plans.

Workforce

Schneider data: US CPG and life-science manufacturers lead global AI adoption

New Schneider Electric data shows U.S. consumer packaged goods and life sciences manufacturers are adopting AI faster than their global peers.

Vendor-sourced adoption data: a useful benchmark if you are in CPG or life sciences, not proof of returns.

Workforce

When a model update resets operator trust, do you stop automation to re-validate?

AWS's Sanjay Mahato and Grayhawk's Tauseef Ahmed answer four audience questions on consistency at scale, graduated trust management and sensor data validation. The central question: when a model update resets the trust operators have built, do you interrupt automation to re-validate?

Treat every model update as a change that can reset operator trust, and decide in advance when it triggers re-validation.

Agents

AI agents in OT: the likeliest failure is an authorized system making a bad decision

Skkynet's Gary Tillery and EverLine's Matt Morris argue boards make OT risk decisions they do not know they are making. AI agents change behavior with every model upgrade and carry no personal accountability, and the likeliest AI failure in OT involves no adversary.

Plan for failure by an authorized agent, not just an attacker, and name an owner for each agent's decisions.

Agents

Schneider, Nvidia and HPE bring AI agents to the plant edge

Schneider Electric, Nvidia and HPE are collaborating to bring AI agents to process plants, targeting OT monitoring, cybersecurity and energy optimization. The agents run on premises rather than in the cloud.

A vendor launch with no reported plant results; the on-premises design is the part to evaluate.

ForgeShift Insights goes out weekly across three editions.

Subscribe on the insights page

We use only essential cookies. We don't use advertising or tracking cookies. See our Privacy Policy.