| Airborne/ GPS Integration: Scrammbled Signals
An airborne GPS project near Las Vegas succeeds in spite of signal interference.
By Ruth Carapella When Aerial Data Service's flight crew left Oklahoma for Clark County, Nevada, they had everything lined up for a perfect airborne GPS mission. Pilot Ron Rinas knew he could count on ADS's twin engine Cessna 310Q. Camera operator Ed Seaton and ground crewman Kyle Gouker knew they could count on the company's Zeiss RMK TOP 15 camera and Trimble GPS receivers. They had flown many successful missions over the last three years with this system.
At the company's Tulsa, Okla., office, ADS had done a thorough job of planning the mission. ADS even knew of a secret military installation north of Las Vegas. But, Rinas and his coworkers didn't suspect that the military installation might be jamming signals, causing ADS's Trimble receivers to repeatedly loose their lock on the satellites. Prework
ADS often handles projects from planning to delivery of a final product. But for this project, ADS served as a subcontractor, providing diapositives, the solutions to aerotriangulation calculations, and the aerial film to Nevada Aerial Mapping of Reno. Rinas said, "With our airborne GPS system, we could guarantee the accuracy they needed and we could reduce the amount of field work considerably."
Rinas wasn't exaggerating. With the company's airborne GPS system, they only required 30 ground targets. Without airborne GPS, they would have needed approximately 250 ground targets to produce a high-quality digital orthophoto of the Las Vegas Valley. It doesn't take a calculator to figure that an 88 percent reduction in the number of ground targets equated to a big reduction in the total project cost.
Before selecting locations for the 30 ground targets, ADS delineated the 30 by 32 mile project area on 7 1/2 minute quad maps. They digitized the flight line locations and built a data base containing the coordinates for the start and end points of the 21 flight lines needed for stereo coverage. They also determined the coordinates for the points where flight lines would cross section lines. The entire data base was loaded into a laptop computer. For the flight, Rinas connected the laptop to the RMK TOP 15 camera and to a Trimble 2000 receiver mounted in the Cessna's instrument panel.
For this project, ground crews in Nevada used Trimble 4000SSE receivers to survey and install the targets. The survey crew determined each target's horizontal and vertical coordinates. They used a steel pin or a brass cap mounted in concrete to mark each target's center. Then they stretched a strip of white target material 2 feet wide and 30 feet long across the target to make it visible from the air.
Finally, ground crews located two tri-stations with known horizontal and vertical positions to serve as base stations at the north and south sides of the project area. Gouker and Will Stout, a representative from Pentacore of Las Vegas, staffed these stations when the Cessna was in the air.
At the base stations and in the aircraft, dual frequency Trimble 4000SSE P-code receivers observed several satellites simultaneously. By using post-processed differential GPS this way, ADS eliminated clock errors and errors introduced by routine military degradation (dithering) of satellite signals. Equipment
ADS attributes a large part of their success with airborne GPS to the Zeiss RMK TOP 15 camera they purchased three years ago. Their RMK TOP 15 was one of the first purchased in the United States.
Jean Carter, ADS president said, "Although it was a huge investment, it enabled ADS to climb to another level of photography. We are now being contacted by clients all over the country because of our expertise in airborne GPS. We couldn't have done it without the RMK 15, our staff's tenacity, or our R & D program." ADS's researchers developed many of the techniques the company used on the Nevada aerotriangulation project.
The excellent optical characteristics of the camera's Pleogon A3 wide-angle lens allow quality image capture under a wide range of lighting conditions. The lens has a focal length of 153 mm and a maximum aperture of f/4. Four internal filters are integrated into the optical system. A tilting device on the camera permits the camera's platform to be tilted upward, providing access to the front of the lens to change external filters.
A forward motion compensation (FMC) device eliminates image motion created by the Cessna's 200-knot speed. The camera's T-AS gyro stabilized suspension mount, the first in the United States, compensates for vibration and rotary movements of the aircraft. Three gyros in the mount continuously measure and transfer correction data to control the camera's position. Using this data, drive assemblies move the camera frame to compensate for the aircraft's rotational movements.
When coupled with Zeiss's T-TL terminal, the flight report includes the pitch, roll, and drift angles for each image. This allows continuous checking of image quality and creates a digital record of the exact camera orientation at the instant of exposure. By knowing the exact camera orientation, ADS can refine their aerotriangulation calculations to eliminate potential errors introduced by the changes of the camera's position relative to the GPS antenna.
The Cessna's GPS antenna receives signals for a Trimble 4000SSE P-code receiver and for the Trimble 2000 located in the Cessna's instrument panel. The 2000 receives information on the plane's location every second and it computes where the plane will be 100 times per second. The laptop compares the anticipated location against the locations established in the preflight database. When the two coincide, the laptop sends a firing command to the RMK TOP 15.
At the midpoint of the shutter open time, the camera shutter emits a pulse that triggers the Trimble 4000 receiver. The 4000 records the exact coordinates of the camera's position, making aerotriangulation calculations possible. Besides recording the antenna's position when it fires, the GPS receiver records the antenna's position once a second. The 4000 can store over four hours of data in its memory while observing six satellites at a one second sample rate.
The coordinates for each image differ from the coordinates established in the preflight data base. Air turbulence and slight variations in the Cessna's 200-knot speed cause the actual point of image capture to vary from the intended point of capture by as much as 10 to 200 feet. The Mission
The project became active in mid April 1996. Chances for clear weather seemed promising, and satellite coverage was good.
On the first morning in Nevada, Gouker and Stout headed out to the base stations. They needed time to get set up over the tri-stations before the Cessna left the ground. But it didn't leave the ground that day; there were too many clouds.
The following morning, with the base stations up and running, Rinas and Seaton ran a 20 minute GPS fast-static survey session on the ground. Everything looked good. Seaton switched the Trimble 4000 receiver into the "rove" mode and manned the camera.
The 4000 lost it's lock on the satellites while the Cessna taxied out. At this point, Rinas and Seaton decided to fly the project using on-the-fly (OTF) initializing when post processing GPS data. The airplane quickly reached its flight altitude of 12,000 to 13,000 feet, an elevation 10,000 feet above the ground surface.
At first, the system seemed to be working smoothly. However, Seaton soon began to have trouble keeping the 4000 locked on the satellite signals. In the first hour, the 4000 lost it's lock 31 times. Each time the receiver missed the signal for a split second it would take the GPSURVEY software two to three minutes to reinitialize. Although they had good photo coverage, Rinas and Seaton didn't have enough GPS data.
Clearly something was disturbing the signals; the company had never encountered this type of signal locking problem before. Eventually, Rinas attributed the problem to the military area north of Las Vegas. He said, "There is a large military installation north of Las Vegas. They were probably doing some kind of testing. I'm sure they were jamming the satellite signals."
To further complicate matters, the shutter on the camera began to malfunction. No one had anticipated this problem. It had never failed before.
Zeiss rounded up a replacement camera so ADS could continue the mission while their camera was being repaired. "They didn't have to do that," Rinas said. "They really got us out of a jam."
Four days later, with the replacement camera installed in the Cessna, Rinas and Seaton tried it again. Once more they started with a fast-static survey session on the ground and once more they lost the lock on the signal while the plane taxied out. They did manage to complete 16 flight lines that day. Seaton had to continually monitor the 4000 to make sure it collected good GPS data.
Seaton also monitored a video camera mounted in the RMK TOP 15 camera body. The video camera allowed Seaton to gain a camera perspective of the photo mission. Drift and photography overlap control data were superimposed over the video image, giving him the opportunity to ensure image quality.
They completed the mission the following day, but not without additional signal problems. They flew six of the flight lines twice and one line four times. Some days just aren't easy! Rinas said, "We'd get on the flight line and two or three miles from the end the signal would quit. We'd have to turn around and refly the whole line because we needed to get the line in a continuous pass."
Rinas limited the Cessna's air time to about four hours per day. When they returned to the ground, Seaton would begin downloading data from the 4000 into the computer to calculate the data for the exact camera exposures. The calculations had to be completed each evening so Rinas and Seaton would know what to fly the following day. Rinas said, "Sometimes Eddy would be up 'til midnight making sure of what we did and didn't get."
In the end, ADS captured 438 frames of data along 34 flight lines to ensure they had good usable GPS data. Their aerotriangulation solution was better than specified, with RMSE values of .327 feet in X, .334 feet in Y, and .417 in Z. Rinas, Seaton, and Gouker can chalk up one more successful mission for ADS. About the Author:
Ruth Carapella writes about forestry and natural resources for Pen Craft Writing & Editing Services in Harrison, Idaho. She may be reached at 800-689-9235 by phone or fax. You can send her e-mail at: [email protected] Back |
