Unmanned “Unplugged”

Jack Langelaan, Penn State University

Jack Langelaan is an associate professor in the Aerospace Engineering Department at Penn State and works with the university’s Air Vehicle Intelligence and Autonomy Laboratory. His work focuses on improving performance in unmanned aircraft. His team also won the 2011 NASA Green Flight Challenge.

You are an associate professor of aerospace engineering at Penn State, and lead a team of researchers who are pursuing new and diverse uses for autonomous flying vehicles. Can you tell us a bit about what you’re working on and what you’re trying to find?

One of our main goals is extending range and endurance for small robotic aircraft. For us, “small” means that the aircraft can be hand launched, and this has some major effects on the aircraft. First, payload is limited, so there is a trade-off between the amount of sensors that can be carried and the amount of batteries that can be carried. The aircraft can carry a small, lightweight sensor package and more batteries so it can fly a bit longer but at reduced sensing capability, or it can carry a somewhat larger, heavier sensor package and a smaller battery, increasing sensing capability but reducing flight time. Second, the aerodynamic efficiency of small aircraft is quite a bit worse than larger aircraft. Since the Reynolds number is lower (Reynolds number is a parameter that describes the ratio of inertial forces in a fluid to viscous forces), drag tends to be comparatively higher on small aircraft than on full-sized aircraft. The average flight time for aircraft with a six-foot wingspan is between one and one and a half hours, and we would like to extend that to all day. We could of course wait a few years for battery technology to improve, but with some good flight control and flight planning, we can improve performance now.

Birds like hawks and vultures are about the same size as many small robotic aircraft, and these birds have evolved methods to exploit the energy in the atmosphere. By doing so, they can fly many hours and hundreds of miles without flapping their wings, and we’re trying to do the same with small aircraft. Birds exploit thermals (columns of warm, rising air) and slope or ridge lift (where the wind is deflected upwards by hills and ridges). Human glider pilots exploit both of these and also soar in mountain waves. Some sea birds such as albatrosses are also able to exploit wind shear by dynamic soaring. We are working on enabling soaring flight for small UAS — essentially we’re developing biologically inspired methods of flight control to improve range and endurance of small unmanned aircraft.

One thing that makes your work special is your aim for UAS to more closely mimic the aerial performance of birds. You have said that fulfilling these applications will require developing planes that can think, or at least react, for themselves. How do UAS “think,” and how will making them do so improve the technology?

Robots think using the computer software that we (the people that design the robots) create. In my research we’re focusing on developing the software that can sense and map the regions where energy can be obtained from the atmosphere as well as fly in such a way that this energy can be exploited by the aircraft. This will reduce the amount of fuel (or batteries) that must be carried, allowing small robotic aircraft to fly all day using almost no energy from fuel or batteries.

Photo courtesy Patrick Mansell/Penn State

Can you explain in layman’s terms what “dynamic soaring” means and what it can be used for? 

Dynamic soaring is a flight technique used by albatrosses. It exploits the fact that wind speed changes with altitude: Close to the surface of the ocean, the wind speed is significantly lower than the wind speed 30 feet above the surface. The albatross begins flying a low altitude, and then it swoops upwards, flying into the wind. Because it is flying into a region of increasing headwind, its speed relative to the air increases, and it can trade this increase in airspeed for more altitude. Eventually it turns back downwind and dives back towards the surface, and the process repeats. As a result, the albatross can fly for over a thousand miles without flapping its wings. 

Radio-controlled glider pilots are doing this as well: They fly on the downwind side of a ridge, swooping from the “dead air” below the ridgeline up into the wind above the ridgeline and back down again. There are many, many YouTube videos showing this, and the current speed record is 498 miles per hour — remember, this is with a glider. Every time the glider crosses the wind shear layer, its airspeed increases, and this cycle lasts for as long as the wind continues to blow or as long as the pilot feels like flying.

What about thermals? We understand you are exploring how to make use of them, much like the way hawks and vultures do (no easy feat). Could you explain for us what thermals are, why they’re valuable to your work, and how you hope to make use of them?

A thermal is a column of warm, rising air. It is caused by uneven solar heating of the ground combined with an unstable atmosphere (so that a bit of warmer air that starts to rise will continue to rise). Often a thermal will have a cumulus cloud (a puffy, cotton ball cloud) sitting on top of it, but not always. This depends on the amount of moisture present in the atmosphere. An aircraft or a bird can ride this rising air to gain altitude. In addition to vultures and hawks, human sailplane pilots, as well as hang glider and paraglider pilots, will fly in a thermal to gain altitude. This is called thermal soaring.

We’re working on flying robotic aircraft in thermals so they can stay aloft all day. The first person to demonstrate autonomous thermal soaring was Michael Allen, who used a modified radio-controlled glider. Since Allen’s flights, quite a few researchers have been working on this problem. Our focus has been on developing a flock of UAS that can cooperatively map and exploit thermals and at the same time fulfill the main mission of the flock. We defined a set of tasks that such a flock would have to perform: Exploring and mapping the local area to find thermals, survey some target and exploiting a thermal are three of these tasks. We run a task assignment algorithm that decides which aircraft will perform each task, ensuring that one aircraft is always assigned to the surveying task. Then our flight control algorithms ensure that an aircraft fulfills its assignment. Tasks are assigned based on an aircraft’s current capabilities: One that is at low altitude will give preference to exploiting a thermal, and one that is high will give preference to either exploration or to target surveillance.

You are no stranger to UAS innovation. In 2011, you led an international team that won NASA’s Green Flight Challenge for fuel-efficient small airplanes and demonstrated that electric-powered flight can be practical. The Green Flight Challenge, a NASA Centennial Challenge program, is intended to spur private development of fuel-efficient small airplanes. What drew you to complete in such an event?  What did you learn?

I first heard about the Green Flight Challenge (GFC) in late 2009, and the first thing I did was figure out the energy efficiency of an electric-powered motor glider. I wanted to know if it would be feasible and get a rough idea of what kind of performance we’d need to be successful. The GFC’s minimum qualifying standard was 200 equivalent passenger-miles per gallon (with equivalent defined by the energy content in a gallon of gasoline; 33.4 kilowatts of electricity is equivalent to the energy content of a gallon of gasoline). This is equal to the fuel efficiency of a Toyota Prius and about six times more efficient than a typical general aviation airplane. The back-of-the-envelope calculations showed that an electric-powered sailplane could do this with room to spare, and so I started to think about it in more detail. It quickly became obvious that a lightweight, two-seat, self-launching sailplane like Pipistrel’s Taurus G2 was an ideal platform from which to start. In 2010 I was at Oshkosh and I went looking for the Pipistrel booth, and one of their main engineers happened to be there. The rest is history.

Photo courtesy Patrick Mansell/Penn State

You called this breakthrough a “Lindbergh moment.” Why? 

Until 1927, flying was about delivering mail and about barnstorming. It was really for the adventurer. When Lindbergh flew across the Atlantic Ocean, he showed that flying could have global reach, and two years later KLM (Royal Dutch Airlines) had set up regularly scheduled service from Amsterdam to Jakarta, Indonesia. Lindbergh’s solo flight was really the event that kicked off the global aviation boom that lasts to this day.

Until 2011, electric-powered aircraft were a very interesting curiosity. They could fly, but there was a definite difference between a fast electric aircraft and one that could fly for a reasonable length of time. The choice was really fly fast for about 15 minutes or fly slowly for about an hour. The Green Flight Challenge showed that an electric-powered aircraft could do both, that it could fly at a reasonable speed and for a reasonable length of time. It was really the first time that an electric-powered aircraft had performed a typical general aviation mission. 

It’s been more than two years since the GFC, but there are several entrepreneurs and engineers developing electric-powered general aviation aircraft that will soon be on the market. 

You also said in reference to these fuel efficiency and limited emissions breakthroughs that the future is bright for electronic flight. In five to seven years from now, for example, you say we’ll see planes that can fly 300 miles for about $7 worth of electricity. What other benefits can society expect to see? How do we get here? 

With a lower cost of travel we’ll be able to get around more easily. A big barrier to personal air travel is the cost of operating the aircraft. Reducing the cost of energy or fuel will go a long way to enabling all of us to fly, and I think this will have a huge impact on our society.

Consider how different life would be if we had access to our own aircraft. I suspect the impact would be similar to what happened as we transitioned from the horse to the automobile. Congestion in urban metroplexes could be reduced enormously. With three dimensions to work in, traffic wouldn’t be as bad, so commute time would be reduced. Emergency vehicles will have easier access. Deliveries will be quicker. The list goes on. 

There are a lot of challenges to address, though. Integrating personal aircraft into the national airspace will be difficult; guaranteeing safety will be difficult; determining the appropriate level of cockpit automation is a challenge (and that may depend on the expertise of the aircraft’s operator or pilot); dealing with weather is a challenge (if I fly to work one sunny morning I’ll still want to fly home even though the afternoon is rainy and foggy, or very gusty). And this list doesn’t mention regulatory challenges. So we have a long way to go, but there are a lot of very capable people working on solving these problems.

What about other applications of UAS technology? We hear you are currently working with Penn State meteorology faculty to develop a plane that could gather data in the convective boundary layer, the poorly understood and hard-to-survey region where thunderstorms and tornadoes originate. Can you explain to readers what that entails?

The convective boundary layer is the lowest part of the atmosphere. It’s thickness changes over the course of the day. Depending on the time of year and the weather on a particular day it can reach up to 5,000 feet thick. There is a lot of mixing happening in this layer, and this mixing is how aerosols such as dust, pollen and pollutants get transported up into the atmosphere. It also contains a lot of water vapor. A better understanding of how these aerosols, water vapor, as well as energy get transported up into the atmosphere will teach us a lot about both weather and climate.

As UAS technology continues to develop, how do you foresee it being used by researchers in the future? 

Right now I think the hardest technical challenge is sensing and avoiding traffic. It is very difficult to see other aircraft in time that an avoidance maneuver can be undertaken. The closing speeds can be very high, and developing a sensor that has the right SWaP (size, weight and power) and can still detect other aircraft is difficult. 

As the technology continues to develop I think we’ll see UAVs used for continuous monitoring of environmentally sensitive areas, such as wetlands, and to perform wildlife surveys. Ultimately a UAS is a robot, and we use robots to do tasks that are too dangerous for people to perform or so repetitive and boring that no one wants to do them (or it’s too expensive for a person to do the task). By those criteria (dangerous, dull or dirty), long-duration surveying or environmental monitoring have really become ideal tasks for UAVs. 

What does that future hold for you and UAS? Do you have any other exciting research projects planned? What would you like to do next? 

In addition to our autonomous soaring research, we’re also working on some projects related to autonomous helicopters. We’re working on autonomous shipboard landing systems, so that unmanned rotorcraft can land on ship decks in conditions ranging from calm to stormy, and on multi-lift, where a team of autonomous rotorcraft cooperates to carry a single, large payload that would be too large for one helicopter to carry on its own. As a side project, some of my students are working on a really small autonomous helicopter. It weighs less than 12 grams (about as much as two U.S. quarters), and it will be capable of autonomous flight and vision-based navigation (although the video processing all happens off board).

My ultimate goal is still to have a small UAV autonomously follow a golden eagle as it migrates along the Appalachian Mountains. When we can do that I’ll feel pretty comfortable saying that we’re starting to master autonomous flight.