EAAers Put The Accuracy Back In By BUDD DAVISSON

It's getting a little mind boggling trying to keep up with all that's new in aviation and Oshkosh is guaranteed to square the boggle-factor. We're hardly inside the arches before the new stuff seems to steam roll over us like a tidal wave, challenging us to assimilate it all. Clearly an impossible task. An unfortunate part of this tidal wave of new developments is that the high profile and obvious advances, like new aircraft designs, sometimes cast a shadow that makes it difficult to keep track of the smaller, less obvious developments. Such is the case with the Differential GPS experiments being demonstrated on the field at Oshkosh '93. Not many noticed the GPS notation on the tails of the Cessna 172 and Bonanza parked inside the entry arches. Even fewer, if any, noticed the innocuous looking box mounted on the gable end of the communications building

with what looked like a cellular tele- lenge facing Don Taylor (Cumulus phone antenna coming out of it. Consultants, 5645 Rico Dr., Boca RaNothing hinted at the magnitude of the ton, FL 33487), Jim Cook (Palm Beach experiment they were demonstrating. Avionics, 3950 Airport Rd., Boca RaThe box and the airplanes were im- ton, FL 33431) and Rudy Penteado portant because, if nothing else, they (Genese, Inc., 3300 University Dr., show what happens when a trio of Coral Springs, FL 33065). EAAers are charged with solving a techDon has been EAA's representative nical problem facing aviation. As is to the RTCA (Radio Technical Commitusually the case, they came up with a so- tee for Aeronautics) committee in lution quickly, simply and inexpensively. Washington, DC for the past year or so The particular problem this trio was and one of the things the RTCA task addressing is actually two-fold. The force has been looking at is GPS techfirst is coming up with a newer, less ex- nology, as an approach medium. Rudy pensive, yet accurate system to replace and Don had known each other for the old UHF ILS technology. The sec- some time and Don had flown the movond challenge is, assuming GPS is a ing map software Rudy had developed candidate for ILS type approaches, de- for use in coupling lap top computers veloping a technology that takes Global to existing GPS units. Jim, an avionics Positioning System (GPS) technology expert and electronics manufacturing and refines it so it is accurate enough type, rounded out the trio when, at the for Category One (CAT 1) approaches. exact moment(!) SPORT AVIATION CAT 1 approaches require 200 foot Editor Jack Cox was telling Don he ceilings and 1/2 mile visibility. should get to know Jim Cook, Jim flew Refining GPS and making it work in right past Don's window in his Questhe CAT 1 environment was the chal- tair Venture. Seems like some things

Rudy Penteado and the Hewlett-Packard 95LX computer used as the moving map screen in the Genese system. 56 OCTOBER 1993

are meant to be!

Don had already received approval from the EAA to set up differential GPS demonstrations on Wittman Airport parallel to those being conducted by the FAA and Wisconsin DOT which utilized technology and equipment developed for the purpose by Trimble Navigation. The Wisconsin DOT was

investigating the feasibility of setting up special CAT 1 GPS approaches for some airports under its jurisdiction. Both set ups were aimed at the same solution, only the Taylor/Cook/Penteado approach was typical EAA: Find the latest COTS (Commercial Off The Shelf) technology and solve the problem for the least amount of money. GPS has become the navigational buzz word of the decade and has very effectively swept every other form of navigation, with the possible exception of the whiskey compass, under the carpet. However, the entire GPS concept, being basically a military guidance system, has been forcefully degraded by the government to remove much of the accuracy making it useless for precision approaches. Originally, under optimum conditions, it would give accuracy

down to about half a meter (18 inches) with a few meters possible in the worse case scenario. The military has encrypted the signal in such a way that you need a P-code receiver to get optimum accuracy and they aren't selling those things at surplus stores and aren't likely to. So, about the best a normal GPS unit can deliver is 200-300 feet horizontally and 400-500 feet vertically. There is a way, however, to take the diluted signal and correct it to regain the original accuracy in a specific loca-

tion. This is done by differentially correcting the signal using a known geographical location as a datum point. The technology isn't all that complex and the hardware, although it would have seemed to be something from Buck Rogers a few years back, is

nothing more than the normal microprocessors we've come to accept as part of everything from GPS units to coffee makers. Basically, a point on the ground is accurately surveyed. This gives a

known location in relation to the GPS

satellites. Then, since the signal speed

is known, a time-enroute for the signal from given satellite is computed. Let's

say, using a grossly out of scale exam-

ple, the calculated time is 400 milliseconds door-to-door (you don't

time it with a sweep second hand!). However, when the signal "finally" ar-

rives at our GPS receiver at the

surveyed location, we find it took an agonizingly slow 450 milliseconds. This means it is 50 milliseconds late,

GPS receiver and wireless modem in the Bonanza used in the Oshkosh demonstrations.

so we automatically enter that deviation in the mini-computer software that's hiding inside the GPS unit at that location. This software has been carefully programmed to take the differences from all of the satellites and assemble a sort of electronic compass deviation card for GPS units working at that location. Using the internal deviation card, the signals from the satellites can be differentially corrected so the indications on the GPS read-out are accurate. So, now that we have a way of generating a set of signals that have been corrected for the military degradation, what good are they to us sitting on the ground? They don't do us any good at all sitting on the ground, but if we could get those same signal corrections into a GPS unit that's mounted in an airplane, that airplane would be capable of achieving the original military accuracy of the GPS system. So, utilizing another one of those increasingly common miracles of modern communication, a gizmo that is basically a wireless modem, we datalink the deviation table up to any GPS unit that happens to be equipped with a wireless-modem operating on the same frequency as ours. A modem on board the a i r c r a f t works with an on-board computer/ controller to feed the corrected signals into the aircraft's GPS and, voila, instant accuracy. Let's see Buck Rogers do any better than that! Those are the basic concepts, but putting them into action at a reasonable dollar is what the GPS Trio was aiming at. Their final product is a two part approach which includes a base station that is really nothing more than a special GPS receiver equipped with a wireless modem. The second piece is a matching modem and GPS unit in the airplane, which is dialed into the ground station so it can receive the signals. The airplane also is equipped with a

Rosetta Microsystems encoder. The Rosetta encoder is incorporated in a number of brand name systems and differs from others in that rather than

working strictly in 100 foot incre-

ments it is also capable of the one foot accuracy required for precision approaches. Equally as important, it also has a computer-type serial port. By hard-wiring that port to the GPS in the airplane, a barometrically corrected altitude signal is always being fed directly to the GPS receiver. The barometrically correct encoder signal is useful because it acts like a satellite perched directly above the aircraft at all times. This gives constant read out of accurate vertical information which, when combined with the triangulation of the GPS satellites, gives extremely accurate three-dimensional information to the GPS output device. The output device utilized by the GPS Trio is a commonly available palm sized computer, the Hewlett-Packard 95 LX. The HP 95 has a specially programmed card developed by Rudy that lets it take the navigational signals coming from both the encoder and the GPS and display them as a moving map on the computer's little 2" x 4" display. The entire unit is clipped to the yoke like an approach plate. But that's not all! The moving map only takes half the screen. The other half is a complete HSI that displays all necessary nav data. Then, if that isn't enough, a push of a button replaces the HSI with an RMI or an ADF read out. And some of us still haven't figured out northerly turning errors on the old whiskey compass. Okay, enough theory! When we were sitting in the C-172 on the line at Oshkosh, running simulated programs through the unit, it dawned on us ...

the best way to understand the system

was to fly it. So, Don slid in the left seat and we saddled up on the right for a hands-on demo. The scale of the moving map can be SPORT AVIATION 57

changed through a wide range, from 3/10ths of a mile (the scale is so huge it only shows a section of the runway) to 42 miles, which, at Oshkosh put everything from Green Bay, Fond du Lac and the surrounding area on the map. As soon as we started taxiing, the runway depiction started moving. At the beginning, the map was oriented with North up. However, the instant we exceeded 3 kts., the display changed to a "windshield" type of depiction, where everything is oriented the way the pilot would see it out the windshield.

classic ILS localizer needle, it was a simple matter to use the map as a way

to get us in the ball park, and use the

HSI to fine tune it. The glideslope information is a vertical scale with a moving pointer on the bottom left corner of the HSI and, until it is within range, it is "flagged" as being in-op by a box of stripes that lies over it. From the point of intercepting the ILS to decision height, it was just like flying a normal HSI guided approach, except the HSI was a LCD read-out on a $400 handheld computer rather than a panel mounted gauge costing ten times that amount. All of the display's movements were very, very smooth and didn't have any kind of "toy" jerkiness to them. The GPS-demo Bonanza has the same basic equipment, but it is coupled to the autopilot and the smoothness of the signals really shows. It centers the glideslope and maintains it so smoothly it would show up even the smoothest pilot. It should be mentioned that any kind of approach can be programmed into the software by the computer key pad. It doesn't necessarily have to overlay the ILS, as it did at Oshkosh. Anything about the approach, from the glideslope angle to the designated point of touch down, can be custom tailored. In other words, if a guy wants to set up an approach to his mountain runway that requires a super-steep angle to a touch down right next to the trees, all he has to do is punch in the parameters. The base station (with the case removed). Also, the ground receiver that is correcting the signals doesn't have to be at It's almost uncanny to be taxiing the airport where the approaches are down a taxiway with the runway being flown. In fact, one of the conshowing on the map and moving just cept's strong points is that its the way we see it out the windows. approximate 30 mile range (line of It's even weirder to watch the display sight) can be utilized to design apas we turn off the taxiway to position proaches to any airport within range. and hold on the runway. The map The coordinates of the touch down moves much smoother and in much point and parameters of the approach more real time than I had expected. are all that need to be entered. Also, This is the result of getting position there is no need for front/back course considerations since front course apupdates several times a second. Once in the air, we headed south- proaches can be programmed in for bound, with me watching the moving each end of the runway and then semap. We increased the scale so we lected at will by punching a couple of could see Oshkosh and as far away as buttons. The flexibility in approach deFond du Lac. Our position on the map sign is astounding. The foregoing is why the Wisconsin was indicated by a continually moving pair of crossed lines and we simply DOT was interested in the concept. flew to the intersection where we Using one unit at Oshkosh, they could not only put approaches at all four wanted to go. The ILS approach path to the run- runway ends, but they could also proway was depicted on the screen as a vide Special CAT 1 approaches at large "V" shaped funnel, and our job Appleton and Fond du Lac. Those who want to use the Special was to line up and fly down that funnel. Since we could clearly see where CAT 1 approaches would have to have we were in relationship to the outer the ability to receive the corrected sigmarker, we just drove to the outer nals from the ground station via data link and the frequency of the data link marker and turned inbound. Since the HSI needle functions as a is, right now, one of the big questions 58 OCTOBER 1993

to be settled.

One possibility for a data link frequency is the VOR band (112-117.95 mHz), as well as others in the regular VHF band and FM sub-carrier ranges. However, one of the leading candidates is using transponder frequencies, like 1030 mHz. If a universal frequency is assigned for the purpose, the same airborne receiver could be used in a number of locations. Otherwise, a means for tuning the data link receiver to the local frequency would have to be provided. The project, as approached by Taylor, Cook and Penteado falls somewhere between being an altruistic move to better aviation and a commercial venture. On the one hand the units they have built are proof of concept systems aimed at giving the FAA more input to aid them in making their decisions. Their systems show what can be done using off the shelf goodies to control costs. Cook and Penteado also have a commercial leaning to their endeavors, since they intend to manufacture complete differentiated GPS ground and air systems under the name Genese Navigation Systems. Because of their commercial aspirations, they had a relatively firm handle on what the systems would cost. In round numbers, to set up a complete approach package for an airport, the ground equipment would cost well under $10,000. This includes a special GPS receiver, the software to drive it, a data link unit and a battery back-up to keep the system on-line for 12 hours after a power interruption. This is the system they had running at Oshkosh. Each airplane would have approximately $3,000 worth of stuff on board total. However, some of that equipment may already be installed there, since this cost number includes the right kind of altitude encoder and the GPS unit. If they are already on board, the additional cost comes down. Possibly under $1,000 and most of that is for the GPS antenna, the data link modem and the palm-sized computer ($400 plus $400 program card). What the guys didn't mention in their sales pitch is that once this thing is installed in the airplane, it automatically has the HSI, RMI and ADF displays because this stuff is in the software. If these capabilities are important to the pilot, that f u r t h e r reduces the amount of money he has to invest in navigational gear. A really neat off-shoot of the whole GPS data link thing is that the same data link could be used for a lot of different purposes. For one thing, it could up-link real time weather info. Even more fascinating is its possibili-

U.S.

DEPARTMENT OF DEFENSE NAVSTAR SATELLITES

PERSONAL COMPUTER CONTAINING QLASSNAV PROGRAM

LJ u LJ L.i -J u U IL u) »: 1-1 m ffl OD

£b db ffi CD

s 31 & fliZ2ffi Si £ LUCJ SES cL O

COMPUTER MONITOR SCREEN DISPLAYS THE

GLASSNAV CONTROLLER & DIGITAL TO ANALOG CONVERTER & CONTROLLER SOFTWARE

AIRCRAFT GLASSNAV

SWITCHING RELAYS TO ALLOW SELECTION OF THE CONVENTIONAL VOR/ILS OR GPS TO PHOVlDt INPUTS TO THF HSI AND A/P

AIRCRAFT HSI AND AUTOPILOT

GROUND DIFFERENTIAL STATION AND MONITOR

AIRCRAFT DIFFERENTIAL GPS SYSTEM

Pictorial Description of the "High Tech - Low Cost" EAA Differential GPS Demonstration ties for collision avoidance. If each

aircraft had its own data link capabili-

ties on the same frequency, then each airplane's position and altitude could be displayed on every other such system within range. This would give each pilot his own ATC radar screen to monitor and he'd always know who was around him. Through the software, it would even be possible to tailor the display so only those aircraft within, say, 2,000 feet of your altitude or with threatening headings would be displayed. The possibilities are limited only by the imagination of the computer programmers like Rudy.

So, where do we go from here? It

topilot at all. It will be months before there are TSO'd units available from FAA's court. In the first place, al- commercial manufacturers, which may though they appear to be behind the include Rudy and Jim's company along concept of d i f f e r e n t i a l GPS ap- with the known brand names. But the proaches, they haven't totally thrown units won't be worth much until the their arms around it. They also have a FAA gets into the act and gives everylot of homework to do, like deciding one some firm design parameters to what the data link frequency will be, design to. Without them, the base stasetting up parameters for the base sta- tions can't be put into operation. tions, etc., etc. As it is now, through the good old As everything stands right now, EAA approach of using your head and everything we've said here is experi- what's sitting on the shelf, the GPS looks as if most of the balls are in the

m e n t a l and c a n ' t be applied to a certified airplane in IFR conditions and can't be slaved to a certified au-

Trio of Taylor, Cook and Penteado have given us all something important to get our electronic teeth into. * SPORT AVIATION 59