Attack of the Drones

A few bright aerospace engineers at U.C. are doing a lot to test the sky’s limits.

It wasn’t even as big as the palm of his hand, and the whir of its motor sounded like a pesky mosquito buzzing his ear. But never mind that this flying contraption looked like a Cracker Jack toy on steroids. Steady at his hand-held control, University of Cincinnati freshman pilot Kyle Wilson’s eyes were focused as he maneuvered his drone skyward—until it bumped violently into the ceiling. Then he brought it down slowly for a perfect landing on his backpack.

His grin said it all: Really? I’m getting college credit for this?

Illustration by Army of Trolls
Illustration by Army of Trolls

Welcome to drone class, or specifically AEEM 3077: UAVs for Urban Communities, where some of UC’s brightest underclassmen come face-to-face with the technology, the opportunities, and the unanswered questions of an emerging industry. First things first: nomenclature. Kelly Cohen, who heads UC’s drone research and development program, doesn’t actually call them “drones.” Those who know typically refer to them as UAVs—unmanned aerial vehicles, the common name in your average video game—or UASs—unmanned aerial systems, a more precise description of technology that can carry a lot of other devices and disciplines high into the sky. Cohen, a professor in the Department of Aerospace Engineering and Engineering Mechanics, was a program leader in an Israeli UAV initiative for 15 years before coming to the U.S. permanently in 2005 as a research contractor at the Air Force Academy in Colorado Springs. In 2007, he joined the faculty at UC.

On a warm September morning in the engineering building, with the elevators and AC on the fritz, Cohen tosses his jacket aside, rips off his tie, and plunges enthusiastically into the world of fuzzy logic, intelligent systems, and multi-agent decision making. In Cohen’s future world, the UAV is decidedly not just a cool gadget flying through the air—it is a commercial delivery system, a cartographer, a police detective, a firefighter. Quite possibly a lifesaver, too, in part because it has a bit of what we might call a brain.

It was while at the Air Force Academy in 2002, working on the expanded use of UAVs in modern warfare, that Cohen had his epiphany about the wide world of drone usage. A forest fire was burning in the mountains nearby and it was picking up steam. “The fire started in the valley and everybody said, ‘Oh, don’t worry. It’ll be all right. We have this under control. We know what to do,’ ” he recalled, his voice tinged with sarcasm. But the fire blew up—consuming homes, jumping over fire breaks, and killing two people. “I realized then what a UAV could have done to help.”

When Cohen was hired at UC, it wasn’t to start a drone program. The then-rookie associate professor was brought in to replace a recent retiree and to teach engineering classes on dynamics and controls. But soon he met Manish Kumar, who had been hired to teach classes on mobile robotics and intelligent systems. A native of India with masters and doctorate degrees from Duke University, Kumar had been working for two years with the U.S. Army on robotics, or as he called it “theoretical work.”

The two men hit it off, hatching a plan over lunches and extended coffee breaks. (“We were on campus so it was coffee instead of beer,” says Kumar.) Ideas morphed into plans and plans developed into what is now a steadily growing academic research specialty. With encouragement from UC, Cohen and Kumar have grown their program with money from state and federal sources. Perhaps the best testament to UC’s expanding influence in drone research was a $500,000 competitive grant they won earlier this year from the National Science Foundation. The student population for the class is still small—about 20 in the post-graduate program and 10 undergrads—but they range from a curious sophomore enjoying a break from Homer and Shakespeare to master’s and doctoral candidates deep into research and design. And with Cohen and Kumar as their guides, they are charting a whole new scientific and ethical frontier.

It was fitting that six years after the Colorado inferno that inspired his newfangled career, Cohen’s first venture into the non-military use of UAVs found him back on the fire line, this time through a partnership with the West Virginia Division of Forestry. With forest fires, UAVs—which can fly vertically and sometimes go where fixed-wing aircraft and helicopters can’t—are able to provide critical real-time data. Working with the National Aeronautics and Space Administration, the National Fire Science Foundation, and the West Virginia foresters, Cohen and Kumar are developing a set of protocols to fit specific algorithms with existing fire prediction models.

“UAVs can add to these models what we don’t have now: real time information that will allow us to predict quickly where the fire is going and how fast, so we can allocate our resources accordingly,” Cohen says. Specially outfitted UAVs could assess wind speed and direction and analyze how a fire is growing; use cameras to see through dense smoke; analyze the abundance of combustible fuel on the ground; and protect firefighters by alerting them to changing conditions.

“Oh, I don’t have any doubt that this will change how we fight fires in the future,” Rodger Ozburn, a regional fire specialist for West Virginia told me. “Often, especially in the bigger fires, what we lack is intel. We’re on the ground and sometimes can only see a few feet in front of us.”

Ozburn and UC have been partnering on the use of firefighting UAVs since 2008 through a program known as SIERRA—Surveillance for Intelligent Emergency Response Robotic Aircraft. SIERRA links UAVs with GPS, video, environmental sensors, and fire prediction software to get inside the mind of a fire. This integration of multi-disciplinary software is what turns a UAV into a UAS. Funded in part by Congress, the first tests with a five-pound drone were performed in 2011 and showed great promise. The work continues, using prescribed, or “controlled,” burns. Cohen calls the system “low risk and low cost” and Ozburn agrees. But he still can’t afford a drone of his own. “We haven’t used it in a live fire yet” due in part to budget constraints, he says. “But I have great hope.”

Of course, in dreams begin responsibilities. As drones have proliferated, so have the headaches. Ozburn worries most about UAVs that get in the way of firefighters. In California this summer, firefighters were hampered on more than one occasion by hobbyists who flew their drones into the path of fixed wing aircraft dropping chemicals over the blaze. “It’s a growing problem,” Ozburn said. “All it takes is one report or one sighting and everything is automatically grounded. We can’t risk a collision, so the fire burns. It’s just stupidity, that’s all.”

Cohen is even more expressive. He labels these hobbyists “pirates” recklessly playing with “their toys,” endangering lives, and threatening the future of UAV technology application. He has a point. Hobbyist drones have dropped in uninvited at the U.S. Open tennis tournament, disrupted an air care flight in Dayton, crashed into the Great American Tower in downtown Cincinnati, and landed on the White House lawn. Then there’s the Kentucky man who shot a drone that was hovering over his house and his sun-bathing daughter.

“There are rules for hobbyists but a lot of them are either unaware of them or they don’t care,” says Tom McMahon, vice president for advocacy and public affairs for the Association for Unmanned Vehicle Systems International (AUVSI), which lobbies for rules that they believe will unlock the commercial potential of drones. So far they haven’t had much luck. Back in 2012, Congress ordered the Federal Aviation Administration to establish rules by 2014 for drones weighing less than 55 pounds and to integrate all UASs into the national airspace by September 30 of this year. Both deadlines came and went with little more to show than a new rule requiring exempt drone operators to register with the Department of Transportation.

The delay in finalizing a comprehensive UAV regulatory regime frustrates McMahon, who says the lack of rules has stunted the industry’s growth. “Right now, unless you get an exemption from the FAA, you can’t do anything commercially with a UAS,” he says. “You can’t make any money.” Exemptions have been granted but the process is slow and arduous. Operators, for example, have to have a sport pilot certificate and have cockpit experience. That narrows the field considerably.

McMahon is cautiously optimistic that the new FAA rules—whenever they come—will require only that the operator pass a knowledge test but not need an actual license. His organization has advocated for regulations that would restrict a commercial UAS to daytime flying only up to 500 feet and at a safe distance away from airports and population centers, and require the operator to maintain line-of-sight control with the flying vehicle. With the appropriate rules and regs, AUVSI’s analysis shows that the economic impact on Ohio could lead to $2.1 billion and 2,700 jobs over the next 10 years; in Kentucky, it could be $537 million and 678 jobs.

Paul Orkwis, who chairs UC’s Department of Aerospace Engineering, agrees. “Our job is to help create the synergistic stew that launches this industry,” he told me. “But there have to be rules.” Sporting a light blue tie decorated with a wide of variety of flying objects (fighter jets, the starship Enterprise, and a flying donut that looks a lot like the mother ship in The Martian), Orkwis envisions UC as a recognized leader in linking UAS development with robust commercial enterprises. “We are engineers first,” he emphasized. “But what we need to do with those engineering skills is to help companies use the technology so they can produce something useful and, in turn, create jobs.”

Martin Rucidlo plans to be ready for that day. His warehouse in a Loveland commerce park is abuzz when I stop by in late September. This is the home of the Horsefly, an eight-rotor UAS that one day may do what Santa has done for centuries: Drop your Christmas presents out of the sky and onto your front porch.

It is an imposing machine. Weighing in at 15 pounds, it measures 57 inches across on its diagonal with a metal tray below its wingspan that can carry a 10-pound package. The eight black whirring propellers are capable of producing speeds of up to 50 miles per hour. It’s a bit noisy, at least inside the warehouse, and you feel a stiff breeze when it approaches. The Horsefly is the pride and joy of Workhorse Group, Inc., formerly known as AMP Electric Vehicles. Rucidlo, who serves as president, believes their drone will revolutionize package delivery, especially in rural areas. Right now, it’s experimental but Workhorse, in conjunction with the UC Research Institute, received FAA certification to conduct more extensive testing.

Walking through the warehouse, the first thing you notice is a brown UPS delivery truck on a lift—an E-Gen plug-in electric vehicle with an on-board generator. Workhorse cut its teeth on this technology and UPS has purchased 125 of these vehicles to hit the road next year, primarily in the Sun Belt. The company aims to marry the Horsefly with the E-Gen. The Horsefly fits on a platform mounted on the truck roof. The driver plugs in a GPS coordinate, fits the package into the tray, and launches it. The drone rises to roughly 350 feet and then proceeds to its destination, descending on a straight line to a designated drop-off area. Delivery completed, it heads back to the truck.

Rucidlo says this approach to package delivery makes a lot more sense than having an armada of drones leaving directly from a warehouse, which seems to be what Amazon is contemplating. “The delivery companies will tell you it’s the last mile that’s the most expensive,” Rucidlo says. In rural areas where residences might be a mile off the main road, UAS delivery can save a tremendous amount of time. “I think we’re on the verge of a Kitty Hawk moment,” he adds.

He’s not the only one thinking that. Nick Ernest, founder of Psibernetix, an intelligent system development company based in Liberty Township, has been hired by the Air Force Research Laboratory to create “an artificial intelligence for the research and development of increased autonomous capabilities to allow mixed combat teams of manned and unmanned air fighters to operate in highly contested environments.” Translation: He’s helping the military figure out how to give pilots a little help from the drones. Kind of. Ernest’s PhD research focused on a concept known as fuzzy logic—a system of intelligent software that uses linguistic classifications to enable robotic devices to assess situations and react in ways similar to how the human brain works. “We’ve just about reached our performance ceiling for what humans can do,” says Ernest, who studied under Cohen at UC, “but we haven’t even come close to hitting that ceiling for machines.”

This has already opened up a Pandora’s box in terms of the efficacy and morality of deploying autonomous drones on the battlefield. A 2012 Department of Defense policy currently prohibits it but the Navy recently had some success testing an unarmed autonomous UAV on an aircraft carrier. The idea of militarized drones making kill decisions without human intervention conjures fears of an accidental war started by an errant flying robot, though Ernest notes that autonomous drones have the potential to be safer than those piloted from thousands of miles away because they are more difficult to hack. Despite legitimate (and scary) public concerns, he has little doubt that one day a war may be fought strictly using drones. “Warfare has always changed with technology,” he says, citing the development of weapons from the English longbow to the hydrogen bomb. “You either keep up with the times and win, or you don’t and you lose.”

The idea of an armed chess game in the sky worries a lot of smart people—people like Stephen Hawking, Bill Gates, and Elon Musk, who don’t think Hal’s coup d’etat in 2001: A Space Odyssey is out of the question. Both Cohen and Ernest agree that the drone business needs to become multi-disciplinary. “My next hire isn’t going to be a mathematician or a computer scientist,” Ernest says. “I need a behavior scientist or a psychologist or an anthropologist. We need people who can study and understand how humans not just relate to one another, but also to machines.”

Maybe that’s why Cohen’s drone class at UC attracts such a mix of engineering and liberal arts students. Cohen is an avid marketer and recruiter, eyeing promising third-year students like a college football coach hunting the high schools for a quarterback with an arm. “I ask them if they want to make a difference in the world and you see their eyes light up,” he says. “My goal is to make my students fearless.”

Ernest is one of those recruits. For him, there’s BC (before Cohen) and AD (after drones). “I was an underperforming and timid sophomore when I first met Dr. Cohen,” he says. Today, he’s a CEO of a company blazing a trail in the development of intelligent aerial robotics. “He transformed and inspired me,” Ernest says, “and armed me with the tools and passion to take on the world.” That and an army of drones.

Facebook Comments