Leo Team
Jan 18, 2026
When Professor Michael Beebe's students at North Central State College designed bridges out of popsicle sticks last semester, something unexpected happened. The students who used AI assistance consistently created stronger designs using fewer materials than those who relied on traditional methods alone.
"Leo suggested more efficient designs requiring fewer sticks compared to their traditional designs," Beebe explained in a recent conversation on the Mechanical Intelligence Podcast. The results weren't marginal. Students were genuinely surprised at how AI could offer perspectives they hadn't considered.
From Automotive Engineer to AI Advocate
Beebe's path to teaching wasn't conventional. After years working in automotive engineering, he made the decision to transition to education at North Central State College, where he's since developed a reputation for hands-on, project-based learning. His students don't just learn theory. They build robot chassis, design golf cart bumpers, and compete in electric vehicle design competitions.
This practical approach made him an ideal candidate to test how AI tools might fit into the engineering design process. Rather than treating AI as a threat to traditional engineering education, Beebe saw an opportunity.
"Students today are more familiar with technology than previous generations," he noted. "That makes them receptive to AI tools."
The Popsicle Stick Experiment
The bridge project became an informal case study in AI-assisted engineering. Students were limited to 100 popsicle sticks for their designs. In the first half of the semester, they used traditional engineering methods. In the second half, they incorporated Leo AI into their workflow.
The comparison was telling. When students used Leo for technical Q&A and to explore alternative design approaches, they arrived at solutions that were both more efficient and more creative. The AI didn't replace their engineering judgment. It expanded their options.
"Leo's ability to provide calculations and alternative perspectives encourages students to think outside the box," Beebe said. "They collaborate with AI as a team member, rather than seeing it as a replacement for human skills."
AI as a Team Member, Not a Replacement
This framing matters. One of the biggest fears educators and engineers share about AI in engineering is that it will make human expertise obsolete. Beebe's experience suggests the opposite.
In his classroom, AI functions like an additional colleague in a brainstorming session. It offers suggestions, runs calculations, and provides verified information from engineering references. But the students still make the decisions. They still need to understand the underlying concepts. They still need to evaluate whether the AI's suggestions make sense for their specific constraints.
"A good engineer is one who understands the complete engineering system and knows how to use problem-solving tools effectively," Beebe explained. That includes knowing when to trust an AI suggestion and when to push back on it.
What This Means for the Engineering Design Process
The traditional engineering design process involves identifying problems, researching solutions, developing concepts, building prototypes, and testing results. AI doesn't replace any of these steps. It accelerates them.
When students can quickly access verified technical information, search for existing parts, and get cited calculations, they spend less time on research and more time on actual design work. The engineering design process becomes more iterative because exploring alternatives costs less time.
Beebe sees this as preparation for the workforce his students will enter. Companies are already using AI tools for technical Q&A, part search, and design inspection. Students who graduate without exposure to these tools will be at a disadvantage.
"Engineers should be users of these tools, not just coders," he emphasized. "They need to understand how to ask the right questions."
The Challenges of Adoption
Beebe is honest about the obstacles. Not every educator is ready to incorporate AI into their curriculum. Some worry about academic integrity. Others simply don't know where to start.
His advice: start small. He introduced Leo in one project, in one class, and let the results speak for themselves. When the dean saw what students were producing, interest spread. Other classes have since requested access to the tool.
The key is framing AI as a complement to existing methods, not a replacement. Students still need to learn fundamental engineering principles. They still need hands-on experience with physical materials. AI just gives them another tool in their toolkit.
Looking Ahead
Beebe believes AI tools will eventually become as standard in engineering education as CAD software is today. He draws a parallel to the transition from slide rules to calculators. There was resistance at first, but calculators ultimately let students focus on problem-solving rather than arithmetic.
"The key is to focus on fundamental concepts while teaching students how to use these tools effectively," he said.
For now, he's planning to expand his use of AI in the coming semester, including potentially integrating it into crash testing simulations at a Honda facility. His students, meanwhile, are eager to learn more.
The popsicle stick bridges were just the beginning.
What Engineering Leaders Should Consider
Professor Beebe's experience offers practical insights for engineering educators and leaders thinking about AI adoption:
Start with a defined project. Beebe didn't overhaul his entire curriculum. He introduced AI into one specific project where the results would be measurable.
Frame AI as a tool, not a threat. Students responded well when AI was presented as a team member that could offer alternative perspectives, not as something that would do their thinking for them.
Measure the outcomes. The popsicle stick comparison gave Beebe concrete evidence that AI-assisted designs were more efficient. This data helped build buy-in from administration.
Teach students to question AI outputs. Understanding when to accept and when to challenge AI suggestions is a skill in itself. Students need to develop judgment alongside technical capability.
FAQ
Does using AI in engineering education compromise learning outcomes?
Professor Beebe's popsicle stick bridge project suggests the opposite. Students still needed to understand underlying engineering principles, evaluate whether AI suggestions made sense for their constraints, and make final design decisions themselves. The AI accelerated research and calculations, allowing students to spend more time on actual design work. Academic integrity concerns are valid, but they're about implementation, not whether AI belongs in the classroom at all.
How should engineering educators start implementing AI in their curriculum?
Beebe's approach was deliberate and measured. He introduced AI in one specific project in one class rather than overhauling his entire curriculum. This created concrete, measurable outcomes that helped build administrative buy-in. Starting small, measuring results, and letting success spread naturally is more sustainable than attempting comprehensive curriculum transformation immediately.
Will AI tools replace engineering jobs or make human expertise obsolete?
Beebe frames AI as a team member in a brainstorming session, not as a replacement for engineering judgment. A good engineer needs to understand complete engineering systems and know how to use problem-solving tools effectively, including knowing when to trust AI suggestions and when to challenge them. The engineers who will be most valuable are those who can collaborate effectively with AI, not those who resist it.
What's the connection between AI in education and AI in professional engineering?
Companies are already using AI tools for technical Q&A, part search, and design inspection. Engineers who graduate without practical exposure to these tools will enter a workforce where their peers already know how to use them. Beebe sees this as similar to the transition from slide rules to calculators or the eventual standardization of CAD in engineering education.
How do you teach students to question AI outputs rather than blindly accept them?
This becomes a core skill in AI-assisted engineering. Beebe emphasizes that students need to develop judgment alongside technical capability. They should ask whether AI suggestions align with their project constraints, whether the underlying math makes sense, and whether the recommendations have been properly verified. Students who can do this effectively will have a significant professional advantage.
Can this approach work for all types of engineering education, or just project-based programs?
Beebe's experience is with hands-on, project-based learning in automotive engineering and robotics. Other disciplines and teaching styles may have different integration points. The principle of starting small and measuring outcomes, however, applies broadly. Engineering educators in other specialties should identify specific projects where AI tools could provide measurable value rather than attempting wholesale curriculum changes.
What should parents and students know about AI integration in engineering programs?
Rather than being a shortcut that weakens learning, AI integration gives students exposure to tools they'll encounter in professional work. Programs that thoughtfully incorporate AI prepare students better for the engineering workforce than those that avoid it entirely. The key is ensuring that AI is presented as a complement to fundamental engineering principles and hands-on experience, not as a replacement for either.
