CUAIR takes first in autonomous aircraft competition.
The judges at this year’s Student Unmanned Air Systems Competition certainly had a flair for the dramatic. Each year, the three-day competition ends with an awards dinner attended by all of the teams, parents, and judges. This year, that meant 35 teams sitting around tables at the Patuxent Naval Air Station in Maryland enjoying a buffet dinner, and waiting to be called to the podium for their awards. The judges started with the team that came in thirty-fifth. Then they called up the team that came in thirty-fourth. And so on. And so on. And so on.
Eventually, there were two teams that had not yet been honored: North Carolina State and Cornell Engineering’s CUAir team. Rather than put an end to the tension, the judge at the podium took a few moments to give a short history lesson. He recalled the very first competition in 2003. Just two teams entered that year: North Carolina State and Cornell. NC State’s plane was unable to fly autonomously. Cornell’s plane flew—right into the ground. Despite the crash, Cornell was named the winner of that first competition.
After recounting the first place crash of CUAir in 2003, the judge went on to announce that the second place finisher in 2013 was…North Carolina State. As the words were leaving his mouth, members of CUAir knew that they were the winners. “It was kind of a nail-biter,” says electrical sub-team leader Joel Heck ’14 EE. “It was very exciting.”
The win came as vindication for Phill Tischler ’13 CS. Under his leadership, the team decided to ditch the kit-built planes they had been using and instead design and build a custom composite aircraft of their own. “It was a great victory not just for the team members, but also for everyone involved—our sponsors, the Cornell community, and friends and family. Without their support, we would not be where we are today,” he says. “We have been working hard all year to achieve this monumental feat, and I have no doubt the team will only improve in coming years.”
For the Student Unmanned Air Competition, remote control systems are turned off, leaving the model-sized aircraft to autonomously locate, identify, and classify targets on the ground using their own on-board imaging systems. They must connect to a directional Wi-Fi network and obtain mission data.
The competition in Patuxent is organized by the Association for Unmanned Vehicles Systems International. According to its website, this non-profit organization exists “to advance the unmanned and robotics community through education, advocacy, and leadership.” Their sponsored competitions give students opportunities to participate in hands-on robotics activities so they can apply their science, technology, engineering, and math skills outside of the classroom. The association has awarded more than $1.3 million in prize money over the years. For their first-place showing in June, CUAir brought home $7,450.
Back in 2003, the year of their crash victory, Cornell’s team took home a first prize of $3,000. Karl Schulze ’05 CS, a founding member of the CUAir team in 2002, remembers it this way: “The first competition was actually a disaster, it was the first year the competition was run, and there were only two teams,” he says. “We ran the first test flight as a remote control plane so that we could capture aerial images to tune our vision algorithms. Shortly after takeoff the plane flew through a radar beam they had either forgotten (or not realized they needed) to turn off, and lost all remote control reception. The plane promptly nose dived into the ground, totaling the aircraft. As a result neither team completed the course, and we won for having a better journal paper/technical design and entry.”
Ironically, for a team that creates and flies autonomous vehicles, that first CUAir team was not autonomous. It was a branch of the Cornell team competing in the association’s autonomous submarine competition. As a high school student, Schulze had competed against Cornell’s underwater team and so he joined them when he got to Ithaca as a freshman. When the association launched the Unmanned Air Systems competition that same year, Schulze convinced team captain James Buescher ’03 ME and team adviser Kevin Kornegay to launch an unmanned air vehicle for the competition. “The first year was really run as CU Autonomous Underwater Vehicle team running two robots, one in the water and the other in the air,” he says.
After winning the inaugural event in 2003, CUAir suffered through an eight-year drought, not once finishing in the top three. Tischler was on the team in some of the lean years and he was part of the group that finally put everything together. In 2012 the Aeolus II, a plane Tischler worked on as the software team leader, took first in mission and second overall at Patuxent. Then, by taking the systems that had performed so well in Aeolus II and putting them in a self-built plane, named Hyperion, CUAir finally managed to reclaim the crown in 2013. Hyperion found all ten targets—four better than the next best entrant, was the only system to image a target located off to the side in the no-fly zone, and was one of only a handful of vehicles to connect to the Wi-Fi signal successfully.
Despite the intense competition for talented undergraduates among Cornell Engineering’s 20 project teams, CUAir has more than 30 members this year. All the teams give young engineers meaningful experience as part of a collaborative effort to address and solve tough engineering problems. Student teams work to provide clean water for communities in the Global South; to design, test, and build a Mars Rover; to design, build, and test genetically engineered machines; and to design, build, and race concrete canoes, among other challenges. In the teams, students take leadership roles and learn both basic and high level engineering concepts and skills. They also learn about effective communication, collaboration, and compromise.
To join CUAir, students apply, interview, and then wait to hear if they have made the cut. Those who are accepted join one of five sub-teams; Airframe, Software, Electrical, Autopilot, or Business. In practice, the members of CUAir take a systems approach to the design of their aircraft since all systems must work together to accomplish the mission. Toward this end, mechanical engineers, aerospace engineers, electrical engineers, computer engineers, and chemical engineers all work together.
Heck is in his fourth year with CUAir. “I applied to join CUAir because at the time it was one of the smaller teams. I had high interest in aviation and aircraft, so my decision was driven by my interest,” he says. “You can take your project team work and get class credit for it, but I don’t. I do it because I love it.”
Heck says that this year’s plane “will look similar to last year’s, but we want to make it fifteen pounds lighter.” He is a firm believer in learning by doing, so in the redesign of the airframe he has separated the plane into its constituent parts. “Last year the airframe was mostly designed by a senior for a senior design project. Other airframe members were involved, especially with the fuselage,” he says. “This year the design has been broken up and assigned to sophomores and juniors. One team is working on the wings, others are working on the body and the tail. It’s a great way to get younger team members involved with some real responsibility.
“Rather than toss out the winning design and start from scratch, says Heck, “We will make some tweaks this year. Our goal is to improve our infrastructure and our design process.”
Members of CUAir don’t learn just from each other. Both Heck and Tischler value the opportunity the competition in Maryland gives them to see the designs other teams come up with. “Even though we came in first this year, NC State really had a great design and we had a lot of questions for them,” says Heck. Tischler says that as team lead he encouraged team members to get out and talk to the other teams to learn from them as well as to share some of what CUAir has learned over the years.
Heck has some clear ideas about where CUAir will go next, but he is a bit reluctant to tell other teams just what is coming. “Over the past 11 years we have continually been building up and we have finally understood the winning formula,” he says. “So now it’s a matter of doing it. We will experiment with other things off to the side and then bring them in as we perfect them. The future looks good.”