Key Takeaways:
- The best AI strategies start with clearly defined problems, not shiny new technology.
- Empowering engineers means blending hands-on learning with a culture that encourages experimentation.
- Early AI successes often come from augmenting routine tasks, freeing time for creativity.
- Leadership alignment is key, but sustained investment and patience matter even more.
In this episode of Great Question: A Manufacturing Podcast, I talk with David Brown, vice president and CTO of TE Connectivity's Transportation Solutions business. From generative design and predictive maintenance to workforce upskilling, Brown explains how TE is harnessing AI to boost innovation and efficiency on the factory floor. The conversation delves into how TE developed its "Better Products Faster" and "Better Customer Service" AI initiatives, created hands-on training for engineers, and why the most significant lesson is that AI works best when people know exactly how to utilize it.
Below is an edited excerpt from the podcast:
LD: Pleasure to have you. So, let's start off with a big-picture view of how AI is currently transforming product development and factory operations within the transportation sector.
DB: Yeah. Let me split that into two, Laura. I'm going to address them separately because we've got initiatives running in both parts of our business. So first of all, if I start with engineering, we've put our AI engineering initiatives under the banner of “Better Products Faster”. So we're really looking at how AI technology can help us get our product development lifecycle reduced and bring better-performing products to our customers. And we came up with a strategy a couple of years ago, and we're well underway in the implementation of that. So just breaking down our engineering strategy, first of all, we broke it into three pieces.
First of all, the establishment of what we call an AI hub, which is a bunch of AI specialists who we've set up, who are serving all of our engineering organization with their expertise. Secondly, we've developed and deployed a company-wide AI training program. So that's available to all of our engineers. And then lastly, we've rolled out our Generative AI solution, which uses our proprietary TE technology, and we call that product “Tell Me”, so that's now available to all of our engineering community. So really, on the engineering side, there are three very separate pillars that work very nicely together to bring AI capability to our wider engineering organization.
If I flip over to the manufacturing side of our business, that's under the banner of “Better Customer Service”. And what we've been focused on here for a little bit longer than engineering is really, again, how can we bring AI to life across our manufacturing footprint? And similar to our engineering, we've got three major pillars here.
The first one is active analytics. How can programs like predictive maintenance make a difference in our operations? With predictive maintenance, for example, we're targeting a 25% reduction in unplanned maintenance. Secondly, augmented process automation. So, for us, that covers everything from cobots across our manufacturing footprint to AMRs (autonomous mobile robots), which are used for material handling across our various manufacturing locations. And then thirdly, and similar to our engineering strategy, we have a generative AI component to our operations and manufacturing strategy, again looking at how Generative AI can do things like improve decision making, improve our inventory handling, improve logistics, etc. So, two different but connected initiatives across those two different parts of our company.
LD: Yeah, that's a very robust initiative you guys have. I don't hear of many companies having quite a thought-out plan around AI. So that's very interesting.
DB: It's something that we have to be agile and willing to change. Certainly ,one thing that's really different about AI compared to other initiatives is the end goal is typically very clear when we start an initiative in the company. With AI—week by week, month by month, the capability of AI is changing. So it's taught us that as we look at AI, we need to be nimble and we need to be very agile.
LD: Yeah, it's rapidly changing across any industry. So you do have to be very flexible. I want to touch on a survey you guys recently did: The 2025 Industrial Technology Index. You guys found that 42% of organizations globally lack formal AI training, though 71% of engineers who responded say they're interested in said training. Why do you think there's a gap that persists despite that high interest from engineers?
DB: Yeah, this comes from our Industrial Technology Index, which is incredibly insightful and we get some really great information from our respondents. Something else that came back from our respondents was that 81% of the engineers who responded said that they think that AI can help them solve complex problems that they face in their jobs every day. So there's certainly a very keen appetite to use AI and to understand how to use AI. But yes, going back to the start of the question, 42% of organizations do lack formal AI training.
Look, I think there are a couple of things here. Number one, as we've worked across our organization, I think where we've spotted challenges is when we're not focused on clear outcomes. AI is two letters, but it encompasses so much. It’s so large. And I think we've got to be very focused on the outcomes for the individuals and the outcomes for the organization. That's something that we've been spending more time working on recently. I think the second one is simply the wrong language, and what I mean by that is not only should we think about the outcome, but also depending on your target audience, how you use AI and where you use AI needs to be different. And I think there’s a little bit of a little bit of fear around “Is there something that's going to be accessible and relevant to me?” So I think getting it right and focusing on those outcomes is key to improving the effectiveness of training for AI.