Text Size
NASA - Perfecting the Autopilot
December 7, 2007


Benjamin was 4, and like any other father, Mark Motter sought a link with his son, something they could do together.

The answer was in the air.

"My father flew C-47s in World War II, and my son and I put together a model of the airplane," Motter says.

The non-flying C-47 led to a meeting with a radio-control airplane devotee, and that led to the Motters building a flying airplane.

Seven years later, Mark Motter is still flying models, but now as part of the Electronic Systems Branch at NASA Langley Research Center. His specialty is re-programmable, or adaptive, control devices, and his airplanes are test beds for increasingly capable autopilots that could well be a large part of aviation's future.

"A typical autopilot is only designed to have so much computing power," Motter says. "But these days, it's quite easy to have excess computing power."

It's something he has learned in experiments that evolved from a 2002 NASA Langley Creativity and Innovation Program, a commercial autopilot and a Styrofoam, 1/9 scale-model Mig 27 donated by he Army's Fort Eustis.

Now Motter and others work with a model something like the Navy's F-14 Tomcat, complete with a jet engine that can power the craft up to 200 mph.

But he bridles at the notion that he's still playing with toys. Rather, the scale-model airplanes are guided by autopilots that record flight data in an onboard memory and seem to get smaller with each passing day. The planes also carry telemetry that sends data to the ground control station, where it is displayed and monitored by James W. High, also of the Electronic Systems Branch, in experiments that are part of the quest for better and more reliable unmanned aerial vehicles for commercial use.

"We've done any number of fully automated demos where, basically, the plane is sitting on the ground and we push a button and it takes off, flies around and does whatever it's going to do and lands," Motter says.

The mechanics for doing all of that, including setting Global Positioning System parameters for a blocked-out-course, are programmed into an onboard computer not much bigger than a credit card. A laptop computer on the ground has a green "takeoff" button on its screen, waiting for the cursor and a click to send the plane into the air.

Besides the autopilot and telemetry, the airplane carries radio-control hardware linked to a device on the ground that's monitored by a technician, often Motter himself. The plane is never out of sight of people on the ground, and the technician is poised to serve as a safety net for the whole process.

"We watch it like a hawk," Motter says. "I've got my finger on the switch, and if it so much as hiccups and I don't know why, I'm taking over."

That scenario reminds that the real stumbling block in advancing commercial passenger auto-piloted flight goes well beyond the ones and zeroes of a computer. When you pay $400 for an airplane ticket, you want the security of seeing a pilot, even if he or she just says "have a nice day." The notion that today's commercial airplanes are largely flown by computer does not negate the comfort a passenger feels in knowing that somewhere ahead, a person is ready to take the controls in the unlikely event that the machine fails.

"(Control systems) have been built to work in conjunction with a pilot in the plane, and you still, at the bottom line, have a pair of eyes up there and a brain and somebody who will make a decision," Motter says. "Even if you had all the same sensory inputs (of a human), is the public going to rely on a computer program to make decisions in the air?"

Maybe not, but there are other commercial possibilities for unmanned aerial vehicles, or, as the Federal Aviation Administration calls them, "unmanned aerial systems."

Flying freight or mail is one.

"In our lifetime, we'll see a commercial 747 take off, fly and land without a pilot," says Jeffrey S. Hill, of Langley's Systems Engineering Branch. With the Airborne Sub-Scale Transport Aircraft Research project, the branch has conducted experiments with airliner models. "One of the challenges we at NASA and the FAA face is to see how to do it safely. The business is there. The money is there. What's going to drive it is money, the opportunity to make a profit."

So while Mark Motter continues to experiment with models that can make that vision of profit into prophecy, Benjamin has moved onto other things. At 11, he's into piano and violin. On the rare occasion that he goes flying with Mark, Benjamin's fingers never get close to a propeller.

More often, Motter says, laughing, "Now, I ask my son if he wants to fly, and he says, 'Aw, no, Dad.' "

Image Token: 
Mark Motter with a jet-powered scale model NASA uses to test autopilot systems.
Mark Motter with a jet-powered scale model NASA uses to test autopilot systems.
Image Credit: 
NASA / Sean Smith
Image Token: 
Page Last Updated: November 22nd, 2013
Page Editor: NASA Administrator