Unmanned aircraft flying
weather missions

By Helen Altonn
Star-Bulletin

A pilotless aircraft has been flying off the Big Island the past week making weather observations.

Instead of sitting in a cockpit, pilots are operating the Aerosonde from the library of the National Weather Service in Honolulu.

The small robotic airplane is launched from a car roof rack at Upolu Point and flown via laptop computers.

The plane can fly as long as 36 hours. It made history last year as the first unmanned aircraft and the smallest aircraft to cross the Atlantic.

It also was the longest Atlantic crossing -- 26 hours, 45 minutes on 1.5 gallons of fuel, said Dr. Greg Holland, senior researcher at the Australian Bureau of Meteorology and technical director of the Aerosonde Project.

A chase plane follows the robotic craft for 25 miles offshore in case other aircraft are in the area. Then control shifts from the Upolu Point crew to the computer pilots at the Weather Service, located at the University of Hawaii-Manoa.

They are Maurice Gonella, mechanical engineer with Aerosonde Robotic Aircraft, and Bill Vaglienti, aeronautical engineer with the Washington-based Insitu Group.

Holland and his team have been flying the aircraft off the Big Island for a week to get temperature, humidity and wind data for the Weather Service. The project is expected to end tomorrow.

"It's really an experiment right now to test how well we can do it and what kind of data we can collect," said Jim Weyman, meteorologist in charge of the Honolulu Forecast Office and Central Pacific Hurricane Center.

"As moisture fields approach the Big Island in a tradewind situation, we're able to gather data off the coast aways ... and see how that develops and impacts the Big Island."

There is no other way to get that information for long periods of time, Weyman pointed out. He said the data will be used to compare with Weather Service models.

Holland said the Aerosonde's distance is limited by radio communications "but we're going to satellite and mobile telephone and can operate anywhere we'd like."

The Atlantic flight was completely robotic, with no communications from 20 minutes after launch in Newfoundland until 30 minutes before landing in Scotland, he said.

The University of Washington and the Insitu Group conducted the Atlantic trial flight. The aircraft was packed with computers, a communications radio, a GPS satellite guidance system and meteorological instruments.

Three Aerosondes were lost in the crossing, Holland said. "But one made it -- that's all that really matters."

The robotic aircraft was conceived as an economical means of making meteorological observations in remote areas, Holland said. However, other major uses have emerged for coastal reconnaissance, air-sea rescue and fisheries surveillance, he said.

"It is so convenient, so cheap and so flexible," he said.

Holland said the off-the-shelf cost is $35,000 for one aircraft and the computer program, and it will be cheaper if it's mass-produced.

He said Tad McGeer, Insitu aeronautical engineer, had the original idea for the planes, and the two mapped out the program.

"The first couple months, people couldn't stop laughing," Holland said. "It came completely out of nowhere. There was no precedent."

Now they're having trouble keeping up with requests for their robotic flying operation, he said. About nine aircraft are available to fly with one or two days' notice, and about 10 others can be assembled from boxes, he said.

The aircraft were developed by Environmental Systems and Services Ltd. in Melbourne, Australia, the Australian Bureau of Meteorology and Insitu.

The Australian government and the U.S. Office of Naval Research have been the major financial backers, Holland said. The Taiwanese also provided funding.

He said "a lot of bug-fixing" was done last year in the aircraft and limited operations are being conducted this year. The team was in Tasmania in January, North Carolina in February and Barrow, Alaska, in April. In July they'll be at the Dayton, Ohio, air show.

He expects in three to four years to have routine Aerosonde operations for weather and other observations.

"We may end up like Boeing, with different models to satisfy different requirements," Holland said. "We've proven all the concepts. There is nothing left to prove."