Engineers often rely on AI and advanced tools to solve complex problems, but some challenges still demand lived experience and hands-on design. In this transcript from a recent Mind Over Machine Q&A interview, Design World managing editor Mike Santora talks with Todd Roberts and Owen Kent of medical technology startup ATDev, a company focused on assistive solutions born out of real-world needs. The conversation centers on the gap between what AI can optimize on a screen and what people with disabilities actually face in everyday life, especially when traveling.
Roberts and Kent walk through a particularly stubborn problem, explain why conventional approaches fall short, and discuss how human-centered thinking reshaped their engineering process. Readers can expect insight into how close collaboration with users drives better design, where AI fits into that picture — and where it doesn’t — without losing sight of the practical realities that often get overlooked in theoretical discussions. The full conversation is below.
Mike Santora: Hello everyone, and welcome to another edition of Mind Over Machine, the show where we feature smart people giving smart advice you won’t get from AI. I’m Mike Santora, managing editor for Design World magazine, and this segment starts now.
Today I’m joined by Todd Roberts and Owen Kent of Berkeley‑founded medical technology company ATDev. Todd, Owen, thanks for being here.
Todd Roberts: Thanks so much for your time, Mike.
MS: For those who’ve watched the show before, you know we like to jump right in.
Can either of you share an example from your career where you had to come up with an outside‑the‑box solution—something you couldn’t simply get from AI?
TR: Yeah, absolutely, Mike. It’s a great question, especially with AI everywhere and seeming like it can solve all problems. In reality, there are still a lot of things artificial intelligence can’t do. We see it as an accelerant for solving parts of problems, not the whole thing.
A recent example for us is flying as someone who uses a power wheelchair. Owen uses a power wheelchair every day. These chairs can weigh 400 pounds or more, they’re wide and long, and you simply cannot get them into a modern airliner cabin.
If you’re going to fly, you have to transfer through a series of different seats to get into your airline seat, while your power wheelchair is stowed underneath the plane, where it can be damaged. It’s a mess, and there’s really no good, end‑to‑end solution right now.
What we came up with—really what Owen and his local team in Colorado developed—is a transfer mechanism that lets him very easily and mechanically move out of his power wheelchair while still maintaining his custom-designed seat. He transfers that seat onto a manual wheelchair, which he can use to go from his home to the airport.
At the airport, he uses a custom quick‑connect mechanism to move from the manual wheelchair into an airport‑provided airplane wheelchair, or aisle chair—the very narrow chair that can go down the aisle of the plane. From there, he transfers into his airline seat, with his custom seating system bolted into place, supported by a very lightweight mechanical structure he designed.
When he flies from Colorado to Pittsburgh, where ATDev is now based, the whole process is reversed: from the airplane seat back to the aisle chair, from the aisle chair to the manual wheelchair, and then from the manual chair into a local power wheelchair he has in Pittsburgh.
AI might be able to help design each individual transfer mechanism or suggest improvements to specific components. But what AI cannot do today is fully understand the entire problem: the lived experience of going from your home to the airport, through the plane, and back again as someone using these wheelchairs—and every challenge you hit at each transfer point.
It was that lived experience, plus some quick, practical mechanical design, that allowed us to pull together a solution fast so Owen could travel and work with me as a co‑founder of our medical technology company.
MS: That’s absolutely fascinating. For many able‑bodied engineers, it’s easy to get so focused on components that you don’t fully consider the logistics of how everything comes together in real life.
Even as an able‑bodied traveler, I sometimes struggle just getting myself to the airplane. So it’s crucial that we have people thinking about these issues and creating the kind of technology we actually need.
Owen, is there anything you’d like to add?
Owen Kent: No, that’s a great summary. I think it’s very important work. AI‑based products are great, but they can’t solve every problem.
From the caregiver perspective especially, there’s so much hands‑on involvement and so many subtle, practical details that you only see when you live this every day. That’s where human insight really matters.
MS: Well, thank you again, Owen. And thank you, Todd. We appreciate your time and insight today.
And thanks to all of you for tuning into this video. That’s a wrap for this edition of Mind Over Machine. For more episodes and content like this, follow us on LinkedIn, Twitter, and Instagram, and visit us at designworldonline.com.