The platform allows small UAVs to land, download data and possibly even recharge batteries before returning to the air for further duty. Photo: Nikolaos Vitzilaios
Steve Conyers is in search of a snappy acronym.
When he comes up with just the right blend of initials, it will provide the finishing touch — a memorable moniker — for an ambitious project that has occupied his time as a master’s student in the University of Denver’s Daniel Felix Ritchie School of Engineering (RSECS).
With help from his faculty advisors and from senior computer science whiz and soon-to-be graduate student Tom Hamill, Conyers has designed and built a prototype mobile self-leveling landing platform for UAVs, which is shorthand for unmanned aerial vehicles.
The platform — which may well net Conyers some billing on a patent — was developed under a National Science Foundation grant awarded to the Ritchie School’s Unmanned Systems Research Institute (DU2SRI), directed by Kimon Valavanis, chairman of the electrical and computer engineering department, and Matt Rutherford, assistant professor of computer science. The University shares the $2.3 million funding — aimed at supporting civilian applications for unmanned aerial systems — with California State University at Los Angeles, where faculty and students are working on complementary ventures.
From the moment he first learned about the project, Conyers wanted to share in the challenge. “I knew that this was going to be the perfect thing for me,” he recalls. “I have always been a mechanic, working on my own cars. I really like working on something from scratch and making it the best it can be. And that was really what the landing platform would be.”
The 3-foot-by-3-foot platform allows small UAVs to land, download data and possibly even recharge batteries before returning to the air for further duty. The wheeled device climbs and descends hills and, once it has parked, levels its platform so that UAVs can alight without mishap.
According to Rutherford, the platform extends the range of those UAVs known as vertical takeoff and landing (VTOL) vehicles. “For VTOL aerial vehicles, the biggest challenge is that they can’t stay aloft for very long,” he says, noting that this category includes helicopters and multirotor vehicles.
And that limits their productivity. Say a UAV is tasked with monitoring crops or a pipeline. If it can periodically land on the platform to download urgent data — perhaps information suggesting that an irrigation system has malfunctioned or an oil leak has sprung — it can return to the air quickly to patrol another parcel of farmland or another stretch of pipeline. Just as important, that information can be transmitted quickly to decision makers who can take action. And the platform, meanwhile, can move to a new location, positioning itself for a subsequent landing.
In the best case scenario, all of this happens without too much human intervention. “A goal of the [DU2SRI] is to make devices autonomous,” Rutherford explains. That means replacing remote control with software and sensors. “We want to remove the need to have someone to steer and drive at all times,” Rutherford says.
Enter the hardware/software team of Conyers and Hamill, both of whom were looking for a project to test their know-how.
“From the get-go, it was going to be a mechanically complex project,” Conyers says, noting that “the self-leveling part of it was going to be a big challenge, and to make it compact and easy to carry and still have all the capabilities that we wanted.”
The trial and error associated with that was taxing enough, and Conyers spent some weeks figuring out whether his initial schemes would work. But another challenge unfolded when it came time to source the various parts needed to build the platform. That’s when Conyers came face-to-face with the hurdles facing every engineer with a production deadline.
“I think one of our biggest setbacks [involved] the actual electric linear actuators that move the parts,” Conyers recalls. “We designed everything around these components that claimed to do everything we needed them to do. But when we got out and started using them, we broke every one of them. And these were the beefiest ones [the manufacturer] had. They were able to push the force we needed them to push, but they were plastic and they just broke. We spent quite a lot of time finding some replacement actuators that were quite a bit larger. We tested them before we did the redesign and found that they would be able to do what we needed them to do.”
Once Conyers knew his design was viable, Hamill was brought in to develop the software needed for crucial functions. He came to the project after an independent study with Rutherford allowed him to explore his interest in embedded systems —computer systems dedicated to a specific function within a larger mechanical or electrical system.
“… We have all these embedded systems in everything now — from microwaves to cars,” Hamill says. “This type of programming can be kind of frustrating. What I like about it is that every embedded solution is specific to a problem. Every solution is unique.”
Certainly the platform presented some unique dilemmas. “One of the functions I wrote was to tell the landing platform how fast to go,” he says. He wrote another enabling the device to collect data — relative humidity and degrees Celsius — from a temperature sensor, information that can be essential for efficient operation.
For Hamill, it was especially gratifying to witness his work in action — an outcome not always available to computer science students. “I can see the results of what my software is doing,” he explains. “It’s not just an output on a monitor somewhere.”
The next phase of the project will involve doubling the platform’s size — a task that will engage Conyers as he pursues his PhD and that will test the team’s problem-solving skills.
“One of the things we learned from this one is that the same design is not going to work on a much larger scale,” Rutherford says.
So it’s back to the drawing board — and back to the search for the perfect acronym.