Recent investments in onshoring U.S. manufacturing, combined with immigration crackdowns threatening the workforce, highlight a stark reality for the industry: There aren’t enough workers.

One way that manufacturers expect to address the shortage is via tech, in the form of increased automation for dull, dirty and dangerous tasks. These are the kinds of jobs where technology not only adds efficiency but also improves safety.

The so-called “3-Ds” serve as a useful lens for identifying where automation delivers the most tangible value. Yet digital transformations don’t happen overnight.

The key question for manufacturers, then, is how to digitize effectively. How can you introduce technology both to train human workers and expand their productivity, while also automating work better done by software and robots?

Experimentation

Manufacturing may have a worker shortage, but there's no shortage of technical solutions that promise to extend the capabilities of current staff, or accelerate training and upskilling. That’s the good news.

The bad news is that it’s hard to know what will work best for your organization until you’ve tried it. You don’t want to invest in technology until you know it will work, but you don’t know if it will work until you’ve tested it.

This is where low-cost pilots become invaluable. Generally, the best approach is to start with a problem to solve, identify technology that can help, and test it on a small scale. An aerospace manufacturer looking to conduct a high-volume product run could make the financials work, but only if it found ways to shorten lead times on either welding or stenciling. Ultimately, the company discovered that it could reduce stenciling time by 84% by incorporating augmented reality into the process.

As an added bonus, the AR solution meant it was easier to train people to do that work, resulting in less downtime for the team and greater productivity overall.

Not every pilot has to be so focused, however. Another company had precise inventory-tracking technology throughout its facility to prevent worker downtime. It also had drones flying around the warehouse for deeper inventory management.

The drones’ value add for inventory management was likely marginal. But the opportunity to see drones “in the wild” was hugely valuable for sparking ideas about how they might be used to solve other problems in the warehouse.

The takeaway: Real-world experimentation is an essential part of any digital transformation in the manufacturing space. Experiments force you to pressure-test ideas before you over-invest.

User-Centered Design

While real-world pilots are essential, they shouldn’t be the first time that workers encounter a new technology. You get far better results when user feedback starts early, before you commit resources to building or testing a tool in the field.

Take user dashboards. These can be valuable tools for floor managers, particularly when a shift experiences unplanned downtime. The next shift needs to know exactly where they’re behind, and what the expectations might be to make up the lost productivity.

The pertinent questions here: Are the dashboards visible to managers throughout their shifts? Are they updated in real time? Have managers been trained in interpreting what's on the screen? Can anyone else view the dashboards? Do they understand how their work is reflected?

End users need to be included in the process of building tools like this so they can surface these and other relevant questions before investing too much time in building a dashboard that doesn’t functionally change the manager’s ability to lead the team.

The takeaway: Don’t wait for the “big reveal” to introduce new tech to users. Get feedback early and often to ensure that you’re pouring resources into solutions that are useful in practice as well as in theory.

Impact Over Hype

With each new technology hype cycle, it’s easy to get swept up. Finding a way to use generative artificial intelligence on the floor feels like an easy win for the next board meeting. But there are risks to this approach.

A floor manager was pushing for another machine, convinced that high utilization was the bottleneck. But once the team took some simple cycle-time measurements, they discovered that the machine was idle more often than expected. Buying new equipment would not have fixed the problem. Instead, they adjusted staffing and training, which delivered better throughput at a fraction of the cost.

Another common pitfall is introducing technology in the areas that leadership finds most exciting, rather than where frontline teams are most ready. The more effective approach is to start with early adopters.

When you begin where enthusiasm already exists, you solve one of the hardest problems in any digital transformation: change management. Experiments run faster, feedback is richer, and success builds on itself in a way that creates momentum across the organization.

The takeaway: Think frontlines, not headlines. The most impactful change for your organization won’t always be the buzziest.

For years, we’ve been touting the necessity of digital transformation in manufacturing, as the industry adapts to ever-changing technologies and grapples with a skilled worker shortage. And that reality hasn’t changed. If anything, the reshoring of manufacturing operations only raises the stakes of getting digital transformations right.

Digital transformation isn’t about finding a magic bullet. It’s about making consistent investments that deliver meaningful change for the operators on the floor. That means impactful change is within the reach of any manufacturer with any budget, as long as it’s willing to experiment, incorporate user input and focus on impact throughout the process.

Jason Hehman is the vertical lead for Industry 4.0 and a client partner at TXI.