Very much under construction!

Test Flights with the Mule

The "Mule" is a gasoline-powered remotely-piloted vehicle that is used to test the flight control computer and algorithms developed at UCLA as part of the Autonomous Formation Flight program. The aircraft was designed to mimic the aerodynamics of the solar-powered aircraft. This explains the inverted V tail, and the angled extensions on the wings. Because the solar-powered flyer is so lightly built and has a very high aspect ratio, its wings flex a great deal in flight, and in particular bend upwards. This gives the plane a pronounced dihedral concentrated near the ends of the wing.

The first test flights were held at El Mirage Dry Lake Bed in San Bernardino County, California, about 85 miles from UCLA. The first flights were checkouts of the Mule itself, and a series of flights was then done to provide environmental testing of the flight control computer. There were serious problems with electromagetic interference (EMI), which is just a fancy way of saying that the computer broadcast radio waves that interfered with the operation of the R/C control system. It was necessary to encase the computer in a steel box to block the emissions. When this proved insufficient, the control and power lines running into the computer were filtered to prevent them acting as aerials. After this, it was possible to fly the Mule under R/C control.

Click for details and larger view. The Mule at El Mirage, ready for flight. As configured for the El Mirage tests, the Mule weighed about 130 pounds and flew at 40 to 55 mph.
Click for details and larger
 view. The Mule in flight at El Mirage.

After a number of less-than-spectacular results, including one episode in which the engine crankshaft broke, autonomous flight of the Mule was accomplished in November of 1998. The Mule took off under R/C control, and was turned over to the computer after attaining a safe altitude. The computer maintained straight and level flight, and the operator took over again for landing. The controller for autonomous flight was developed by Laurence Mutuel, then a doctoral student at UCLA.

Over the next year, various improvements were made to the Mule, the most significant being the replacement of the original single-cylinder engine with a two-cylinder version. The single-piston engine caused a strong vibration throughout the airframe, which had to be filtered out of the inertial instrument input, and which led to fatigue ofboth airframe and instrumentation.

It was also decided to change the location of the test flights. While El Mirage has many benefits as a flight range, it also has many drawbacks. Among these are that it is in the desert. Really. We're talking serious desert, with serious heat at noon in July. It was usually necessary to get to the lakebed before dawn, set up, and try to get off the desert floor by 11:00 am. Another effect of being in the desert is total lack of power or water. And, El Mirage is a habitat of the endangered desert tortoise, which means great care must be taken not to disturb the land.

The biggest problem, however, occurs in the winter. The operative term in "dry lake bed" is dry. Once it rains, the lakebed is no longer dry, and is unusable (and off-limits) until it dries out again late the next spring.

For a time, flights were performed at Rosamund Airfield. However, this led to difficulties due to ambient radiation from nearby microwave relay stations. So, on the advice of Ron Scaggs, the new Mule operator, we moved our flight operations to McMillan Airfield, on Camp Roberts, which is a former US Army base and currently a base for the California National Guard.

Click for details and larger
view. Refitted mule on the runway at Camp Roberts.
Click for details and larger view. The Mule in flight at Camp Roberts.

The current (February, 2001) effort with the Mule is centered on autonomous formation flight between it and the Frog, a UAV owned and maintained by the Naval Postgraduate School. The Mule will first be brought back into full readiness and will make several autonomous check flights with its new engine. We will then mount a flight control computer on the Frog, so that the Mule will have accurate information as to its position relative to the Frog. The first formation flights will have the Frog flown by and R/C pilot, and the Mule will maintain formation. By summer of 2001, the plan is to have the vehicles flying in formation, both autonomous.

Update, Fall 2001: The Mule bought the farm, in spectacular fashion. Details and pictures are here. The test program continues, with just the one Frog for now, and with plans to add a second to replace the Mule.