AI probably isn’t going to run your CNC machines anytime soon. And honestly, that’s not where I think the biggest opportunity is for most small- and mid-sized manufacturers.
The real opportunity is using AI to eliminate the hours of wasted effort that happen around the machines. There is a tremendous amount of work happening every day that requires people to search for information, compare data, recreate documents, look for patterns, follow up with customers, and figure out why something went wrong. Much of that work is necessary, but it doesn’t necessarily require a highly skilled person to do it manually.
That is where AI can make a difference. Take estimating, for example. Before a quote ever reaches the customer, someone may spend hours reviewing an RFQ, looking for similar jobs, checking specifications, digging through old quotes and trying to determine whether the assumptions being made are reasonable. AI can help review RFQs, identify missing information, compare a new opportunity against historical jobs and flag potential estimating issues before they turn into expensive mistakes.
It can also help identify repeat quoting mistakes. If your company consistently underestimates setup time, misses certain material requirements or makes the same type of pricing error, that information may already exist in your historical data. The problem is that nobody has the time to go looking for the pattern. AI can help find it. That doesn’t replace the estimator. It gives the estimator better information before making the decision.
The same thing applies to tribal knowledge. Every machine shop has people who know things that aren’t written down anywhere. They know which jobs tend to cause problems, which customers have unusual requirements, which materials behave differently, and why a particular process was changed several years ago. That knowledge has tremendous value, but if it only exists in someone’s head, the company is vulnerable every time that person is unavailable or eventually walks out the door.
AI can help organize that knowledge and make it searchable and usable. Instead of asking someone who has been there for 25 years, “How did we handle this last time?” you can potentially find the answer in the company’s accumulated information.
Quality is another area where I see a lot of potential. Manufacturers already collect enormous amounts of information about defects, scrap, rework, nonconformances and corrective actions. The problem isn’t always a lack of data. It’s that the data isn’t being analyzed in a way that makes the underlying patterns obvious.
AI can help summarize quality issues and identify recurring patterns. Are certain defects showing up repeatedly on the same machine? Are they associated with a particular material, process, customer or type of job? Are there issues that keep coming back even after corrective actions have been implemented? The AI doesn’t need to make the quality decision. It can help your quality and production people see what is happening faster so they can make a better decision.
The same principle applies to your ERP data. Most manufacturers have years of information sitting inside their systems about jobs, labor, machine performance, scrap, downtime, scheduling and production. Yet having all that data doesn’t automatically mean you’re getting useful information from it. AI can help identify bottlenecks, anomalies and patterns that are difficult to see when you’re working from individual reports or spreadsheets.
And this isn’t just theoretical. I’ve seen a shop with nine machines identify approximately $500,000 in defect-related waste reduction opportunities by using AI and the data already being generated by the operation. Think about that for a minute. Nine machines. Approximately $500,000 in potential waste reduction.
The opportunity wasn’t about replacing nine machinists. It was about making hidden waste visible and giving the people responsible for the operation better information to act on.
There are plenty of other practical applications. AI can help draft work instructions, create customer-specific documentation, summarize quality issues, organize production information, make tribal knowledge easier to access and automate sales follow-up that otherwise gets pushed to the bottom of someone’s list.
And that’s where I think manufacturers should start thinking differently about AI. Don’t start by asking, “Where can we use AI?” Start by asking, “Where are our people spending time doing work that doesn’t require their highest-value skills?”
Where are people copying information from one system into another? Where are they spending hours searching through old files? Where are they recreating reports that someone has created dozens of times before? Where do the same mistakes keep happening? Where does important knowledge live exclusively in one person’s head? Where are decisions being made with incomplete information because getting the right information takes too long?
Those are the opportunities worth investigating. Because the biggest opportunity isn’t replacing skilled people. It’s removing the work that prevents skilled people from doing what they’re best at.
Your machinists, estimators, engineers, quality people and salespeople have experience and judgment that you can’t simply automate away. The goal should be to give them better information, eliminate unnecessary work and free up more of their time for the things that actually require their expertise.