By Christi Mays
Downtown Belton traffic may not rival the gridlock of Austin at rush hour, but it only takes sitting a dozen cars deep at 5:15 p.m. (waiting through multiple light cycles) on Main Street for patience to begin wearing thin. It’s the kind of frustration most drivers learn to tolerate. But for senior engineering majors Clayton Johnson and Jesse McCaslin, those routine traffic jams sparked an idea.
By their senior year, roommates Clayton and Jesse had grown frustrated with the traffic on their short drive to campus. What should have been a quick five-minute trek often stretched into a 15-minute crawl during peak times, so they decided to come up with a solution. For their senior capstone project, they teamed up with Joshua Cochran and Noah Baca, to develop an AI-driven approach to improve traffic flow in Belton.
“This was a student-generated idea,” said Dr. Paul Griesemer, associate professor and Engineering Department chair. “They were interested in using AI to optimize traffic light signals and see what kind of improvements in the congestion, in Belton specifically, that they could generate.”
Griesemer knew from the start that the scope of the project was significant, and possibly even beyond what a typical undergraduate capstone might tackle. But the group dove in.
Their first task wasn’t building AI but rather understanding the system they were trying to improve.
The team contacted TxDOT and received an extensive dataset outlining how Belton’s traffic lights currently operate, everything from timing sequences to trigger conditions for light changes.
What they received, in Griesemer’s words, was “an enormous document.” Even after narrowing their focus to downtown Belton and the area surrounding campus, they were still working with more than 50 signalized intersections.
From there, the students turned to Eclipse SUMO, an open-source traffic simulation software used by civil engineers. Over the course of a semester, they painstakingly recreated Belton’s traffic network inside the program, mapping roads, intersections, traffic volumes and signal behaviors.
“They downloaded a map of the city and went through intersection by intersection to define where the traffic lights are and then put the TxDOT information into it,” Griesemer said. “They created a fairly comprehensive simulation of traffic through Belton.”
The result was a fully interactive digital model of the city with cars flowing through intersections in real time, so traffic patterns could be observed, tested and adjusted.
Once the simulation was up and running, the students began the next phase, which was integrating artificial intelligence.
“They figured out how to interface with that simulation, to train AI agents to run the traffic lights in place of the information that they got from TxDOT,” Griesemer said.
In practical terms, that meant teaching a computer system to make real-time decisions about when lights should turn red, yellow or green based on evolving traffic conditions instead of preprogrammed timing.
To explore different possibilities, the team developed two AI systems that were working toward that goal; however, training the AI was no quick process.
“When you're training AI agents, their behavior starts out scattered, all over the place,” Griesemer said. “But over time, they start learning that some strategies work better than others.”
Toward the end of the second semester, as the group continued tracking total wait time across the simulation, they finally saw the numbers begin to shift.
“You could see that metric falling as the AI got better and better and better,” he said, but then the semester ended.
By that point, the system was approaching the efficiency of the current traffic light programming, with the potential to surpass it given more time. “We saw the potential for improved traffic flow,” Griesemer said.
Behind those results was a significant time commitment. During the first semester, each team member logged an estimated nine to 12 hours per week on the project, with some weeks reaching closer to 20. In the second semester, the focus shifted toward monitoring and refining the AI as the systems they built began to take on more of the workload themselves.
As part of their agreement with TxDOT to access traffic data, they shared their findings with the agency, giving a glimpse into what future traffic systems could one day look like.
“There’s always the possibility that it could be implemented, used or improved upon in the future, if somebody takes it over at TxDOT,” Griesemer said.
And even if the project never directly changes a single traffic light in Belton, its impact is already clear in another way.
“At the very least, this team of four students has taken a very, very difficult problem and simulated it,” Griesemer said. “They have applied leading-edge neural net technology to this problem and figured out how to configure everything to make it actually improve the situation. The skills that come with that transfer incredibly well into their future careers as engineers.”
For Clayton and Jesse, who both secured positions with a national construction firm, and for Joshua and Noah who graduate in December and plan to go into mechanical engineering, the project represents more than just a capstone.
“It helped us because we can go into any field and just tackle it from the ground up,” Clayton said. “We can go from nothing to something that works and can be proven to work.”
That thought process is also playing out in a hydroponics project that blends engineering, sustainability and service. Building on work from previous classes, another engineering team is designing a system for Feed My Sheep in Temple, where an unused greenhouse could become both a food source and a job opportunity for women in a nearby shelter. The goal is ambitious: an eventually off-grid, solar-powered system that uses natural nutrients and recirculates water efficiently.
Students like Trystin Brown and Carson Gaido are already deep in the details, turning concepts into something tangible.
“So the buckets we’re using are to primarily grow tomatoes,” Trystin said. “We’re using a technique called deep water culture, which is basically soaking the roots.”
The system expands from there and includes towers for leafy greens, channels that distribute nutrients and carefully designed flow systems to ensure each plant gets what it needs.
“Water will go through the top and then kind of disperse,” Carson explained. “These holes are situated so each one of these plants around the side will get even water and nutrients.”
Many of the ideas for student projects begin in a course called Engineering for Humanity, where students partner with local nonprofits to identify needs and propose solutions. Only a handful of those ideas make the leap into senior capstone projects each year, but when they do, the impact is clear.
“Honestly, a lot of engineers struggle connecting how their work impacts people, because a lot of times it’s so far away from the end user,” Griesemer said. “One of the things that we’re trying to do is give them the experience here so they know engineering can be used to do something that is really going to help people.”
That approach to engineering is a notion that is reflected in all the engineering projects each year, like one that evolved from a thermodynamics class in which students worked to make an engine powered by heat from the parking lot.
“When you put the machine on a hot surface, you can actually watch the flywheel of the engine spin up when it comes in contact with the temperature difference,” Griesemer explained. “Eventually, we're going to put it on a little car, so this thing will be able to just drive around forever.”
For another student-led project, a group of golf enthusiasts set out to improve a common frustration at the driving range.
“This was another student-sourced project that identified a problem in real life and wanted to find a solution,” Griesemer said.
The idea grew from one of the students’ experiences working at driving ranges, where large ball-retrieval machines often struggle to collect balls when the turf is wet. After multiple iterations, the future engineers designed a simpler, user-friendly backpack apparatus that retrieves balls without embedding them in the wet turf.
Griesemer said he loves that all the projects were “so different” and intentionally designed to help students learn by doing, failing, refining and trying again.
“With these projects, it is not like the professors are holding their hands and guiding them. They are taking the initiative. They're learning all about project management and how to organize themselves and optimize their teamwork and things like that along the way,” he said. “We give them feedback when they ask for it. But a lot of these things, students just go out and make it happen.”
He believes hands-on problem-solving helps show students how engineering can make a meaningful impact in society – a perspective that has also shaped his own journey in engineering and education.
After earning a bachelor’s degree in mechanical engineering at Rice University, he completed a doctorate in aerospace engineering from the University of Texas, where he also helped teach engineering students who needed extra support. When he finished school, he worked as an aerospace engineer and later as a consultant on missile trajectory planning and NASA and Air Force projects before returning to academia.
“When I got out in the workforce, I just didn't have that sense of fulfillment in my job. I missed being in the classroom,” he said. “I loved being part of their learning process and seeing those light bulbs come on.”
He describes his path to UMHB as one guided by faith and a willingness to follow where he felt led. When he arrived 15 years ago, the engineering program consisted of a handful of pre-engineering classes, and Griesemer played a key role in launching the full engineering program in 2019 and securing accreditation in 2022.
Griesemer is grateful his journey led him into engineering education, where his heart feels called to be, and that same perspective guides his approach to teaching engineering which he says is more than solving problems, it’s about helping students connect their work to something meaningful.
“A lot of young people want their work to have value and meaning,” he said. “If this is their vocation, they want to connect it to doing good work. The lack of that connection, a lot of times, leaves you doing engineering for engineering’s sake. Just loving equations and models and making things is good for some people, but I think for a lot of people, they want their career to be about more than just numbers.”
###
