The low-touch operating model, made concrete, and the one shift that starts it.
A SafetyChain series, built on the executive roundtable framework from Chris Brandsey, Frame & Flight.
Walk your floor at the start of a shift and watch what your best people actually do for the first hour. A lot of it is not the work you hired them for.
Your shift lead pulls last night's output off one screen. She cross-checks a hold against a quality record in a second system. She calls the warehouse to find out whether the ingredient lot that came in at 4 a.m. cleared incoming inspection, because the COA is sitting in someone's inbox instead of anywhere the line can see it. None of that makes a case of product. It's your most experienced person spending the front of her shift moving information by hand between systems that were never connected.
Once you name that pattern, you see it on every shift. In most plants, people are the integration layer. When your quality system, your production record, and your supplier documentation can't pass information to each other, someone carries it across. Chris Brandsey has a phrase for this: people acting as middleware between systems. It's an accurate description of where a lot of skilled labor quietly goes.
That's the starting point for what Chris calls low-touch operations, his read on where manufacturing is heading after the long move from paper to platforms. Low touch describes a plant where your team spends fewer hours relaying, chasing, and re-keying information, and more hours on the judgment only they can bring. Worth saying plainly: this is about ending the manual movement of information, and not about running the plant with fewer people. Here is what it looks like when it starts to happen.
Less coordination, because the handoffs shrink
Count the handoffs in a single deviation today. Someone spots an out-of-spec reading. They log it, then message a supervisor, who pulls the batch record, who asks quality to weigh in, who asks the operator what they saw, who checks whether the same thing happened last week. Half of that is people confirming what a connected record would already show.
When the record travels with the event, most of those steps disappear. The reading, the batch, the operator who caught it, and the last three times it happened are attached to the same place. Fewer people get pulled in to reconstruct what happened, and the ones who do arrive already knowing the context. You feel it first in the small things: shorter shift handoffs, fewer "can you confirm" calls, less standing around a screen deciding whose number is right.
Faster decisions, because the context comes with them
The reason plant decisions take too long is rarely that the decision is hard. It's that the information needed to make it is scattered. By the time your team has gathered the readings, the spec, the supplier history, and last quarter's trend, the moment to act cleanly has often passed and the choice is now a cleanup.
A low-touch plant collapses that gathering time toward zero. When someone has to decide whether to release, hold, or rework, the context is already assembled: current readings against spec, the relevant history, the supplier lot, who signed off on what. The decision itself still takes judgment, and it always will. What changes is that judgment gets applied to a full picture in minutes instead of a partial one an hour later.
Less time hunting for data
Ask your quality manager how much of audit prep is analysis and how much is retrieval. The honest answer, in most plants, is that the bulk of it is finding records, confirming they're complete, and assembling them into something an auditor can follow.
Less data hunting means the information is where you'd expect it, it's trusted because it was captured at the point of work, and it's ready to use without a reconciliation project first. Your people stop being search engines for their own plant. The record of what happened builds itself as the work gets done, so the question shifts from "can we find it" to "what does it tell us."
Data that moves on its own
The end state Chris describes is autonomous data flow: the right information reaching the right people at the right time, without a person routing it. A supplier COA clears and the receiving team sees it without a phone call. A reading trends toward a limit and the supervisor knows before it breaches. A corrective action closes and the record updates everywhere it needs to at once.
This part can sound like automation. What makes it work is connection: joining the systems that hold your data so the data can travel on its own, without handing any decision to a machine. That connection is the difference between a plant that reacts after something has gone wrong and one that sees it coming.
The shift that starts all of this
None of the four signals starts with AI. Each one starts with getting your people out of the middle of the information flow. That's the shift worth funding first, and it's the one that makes everything after it possible. A plant where people still carry data by hand cannot get faster by adding a smarter tool on top. It just gets a faster way to produce the same fragmented picture.
The paper-to-platforms era digitized how work gets captured. Low touch is the next step, connecting that captured work so it moves on its own. Your best people come out of the integration layer and go back to what they're good at: reading a situation, weighing a tradeoff, and making the call.
Where this leaves your team
The plants furthest along this path did not buy their way there with a single tool. They built a foundation first, one where quality, production, and supplier information live in a connected system instead of separate ones, captured at the point of execution and visible to the people who need it. That foundation is what a platform like SafetyChain is for, and it's the groundwork every later step depends on.
If you want to know how close your plant is to operating this way, start with an honest look at your foundation. Our AI readiness checklist walks you through what to have in place across Chris's four dimensions, data, process, governance, and ownership, and shows you where the gaps are before you invest a dollar in AI.
This series was developed with Chris Brandsey of Frame & Flight, whose executive roundtable framework anchors the ideas throughout.
