Datalink days

September 1, 1995

Wouldn't it be great to get those National Weather Service radar composite images in the cockpit when you really need them? What about hourly observations and forecasts for any reporting station? One more thing I'd add to my cockpit wish list would be some help in spotting traffic en route and at my destination after having successfully negotiated the weather challenges — all at an affordable price. Sound impossible? Well, there may be some relief in sight.

Thanks to some new technology under development by the Massachusetts Institute of Technology's Lincoln Laboratory, in cooperation with the FAA's datalink integrated product team, those of us flying light single-engine aircraft may be able to tap directly into the mother lode of weather and traffic information via datalink. The AOPA Air Safety Foundation is participating in this project because we believe it can address some of the key general aviation safety issues such as weather and collision avoidance. The foundation is under contract to the FAA and MIT to conduct a real-world evaluation of the datalink equipment.

Before dismissing this as just another expensive gizmo that you can't possibly afford, understand that the scientists who are running this project are general aviation pilots who are keenly aware of costs. While no one has built a production model, the demo units are using production hardware; and in quantity, this might be delivered at $5,000 to $6,000.

This summer we flew a 1978 Cessna 172 equipped with an ARNAV Systems multifunction display (MFD), using a variety of pilots. There were high-time instrument pilots, VFR private pilots, students, and CFIs. We wanted to provide FAA and MIT with a real-world look by people who had not grown up with the equipment, didn't know datalink from databank or database, and were not necessarily "high-tech cockpit oriented." The objective was to find out how user-friendly the design was. If the system could be misprogrammed, misunderstood, or distracting, we intended to find out.

The term "datalink" means nothing more than transferring information between the ground and cockpit without voice. It is, in effect, an airborne modem to the ground, to use computer lingo. There are a variety of ways that information can be linked from ground to cockpit. There are radio bands using VHF and FM frequencies, satellite links, and the Mode S transponder. The system chosen for this demonstration happened to be Mode S using a Bendix/King KT70X transponder.

The key to successful datalink is the timeliness of the information relayed, the capacity of the system to handle the traffic, and the cost per message. The ultimate linking device could be any of the above or a combination. The technology is still shaking out, and each has pros and cons. Ground-based datalinks utilizing VHF or Mode S require that the aircraft be within line of sight of a ground station, so many ground stations are required to achieve good coverage. Satellite links have wide coverage but at present are prohibitively expensive.

While we don't have many midair collisions, that's not very comforting if you're the one who failed to see and avoid. Last year there were nine collisions, involving 18 aircraft. The number of close calls isn't well documented but certainly ran much higher than the actual collisions. The "big sky-little airplane" theory no doubt providentially intervened as much as pilot vigilance to keep the statistics low.

For this test program, MIT used a less sophisticated version of the airline Traffic Collision Avoidance System, but it works very well. The traffic information service (TIS) in our 172 graphically shows the bearing, distance, and relative altitude of other aircraft and receives the information from a ground sensor. The sensor gets its input from transponders that ATC sees on radar. Unlike the most sophisticated air-carrier version, the pilot has to decide on any evasive action rather than being directed by the system.

In testing, the preliminary comments were almost overwhelmingly positive. We frequently flew in the typical four miles of milky haze visibility that typifies East Coast summers. The system definitely increased our ability to identify the traffic. MIT found that pilots using TIS could visually acquire targets eight times faster than those who just looked out the windshield, using traditional scan patterns.

While TIS will not yet detect aircraft without transponders, those aircraft represent a small part of the population, particularly in high-density airspace. According to a recent AOPA survey, an estimated 93 percent of all GA aircraft now have transponders.

A Cessna 172 is not what one ordinarily thinks of when the weather turns to cumulonimbus; but as the world's most popular aircraft, it would likely be used for many more cross-country trips if pilots had more confidence in their ability to avoid the weather. The weather function allows both text and graphic radar returns.

Want the latest sequence reports or terminal forecasts for any reporting location in the country? Just punch in the identifier and wait 15 seconds. They'll come back in current "teletype" code, or they could be modified to give you plain English — although that takes more transmission time, which is something to consider.

The graphical weather service (GWS) allows pilots to request radar returns from anywhere in the country. This is done by pushing some buttons on the edge of the multifunction display to select the three-letter identifiers of the airport or VOR where you want the composite radar picture and to request a 25-, 50-, 100-, or 200-mile-diameter view. Most of the pilots handled it fairly well. With a few tweaks to the software and hardware, this very good equipment will become exceptional.

GWS was useful in locating routes to avoid widespread areas of precipitation or in figuring out that it was time to land and let the weather do its thing for the next hour or so. There was no local weather on the day of my first evaluation, so I asked for Miami, 800 miles away. Had we been flying down from the north, an early fuel stop would have been just fine — a solid line of thunderstorms stretched up the coast almost to Daytona Beach. The decision was easy once the picture was displayed.

It takes the system about 15 to 20 seconds to send a request down to the ground sensor, which processes the data and shoots it back up to the aircraft. Since graphics are memory intensive, some compression algorithms are used to speed up the transmission process. This results in some weather returns looking like polygons — not what we typically think of in a radar precipitation display. This didn't prove to be a major problem, since the information is intended for strategic avoidance, much like lightning detection equipment. Currently being tested is a new compression algorithm that should improve the picture considerably.

GWS is really not intended to be used like airborne weather radar, which is sometimes used to pick a route between thunderstorm cells. There are two reasons for this — the compression algorithms distort the radar returns and, more important, the image update cycle can be as much as 15 to 20 minutes old. In a dynamic weather situation, 10 minutes can mean the difference between a tolerable ride and an unforgettable one.

The delay is caused by ground processing that takes the raw Nexrad data and removes ground clutter and anomalous propagation, which make weather returns appear where there is no weather. MIT and WSI, the weather information provider, are working on this; and if they were to cut it in half, that would be most helpful. This is more critical for faster aircraft, for which a 15-minute lag translates into roughly 40 or 50 miles. We experimented with a seven-minute version that doesn't have all of the processing of the 15-minute version, but my feeling is that seven minutes is the way to go when cells are popping all around.

The GWS currently provides only "north up" orientation on the display. This is probably desirable when requesting the weather 300 miles ahead, but some of our evaluators preferred that nearby weather be presented as "course up" to help them visualize it relative to the aircraft. This can be programmed into the system and would require only some input from a GPS or loran receiver.

Despite some drawbacks, GWS has some definite advantages over airborne weather radar. It is a much simpler device with no moving antenna parts and is not subject to attenuation. Attenuation occurs when there is so much rain ahead of the radar that its signal is absorbed beyond a certain distance and the radar becomes "blind." The GWS uses the ground-based Nexrad radars, which are far more powerful than any airborne radar unit — and the pilot can get the "big picture" anywhere in the country.

Was a convective sigmet just issued? I've always had trouble with the "From 20 northwest of Bigwich to 15 southeast of West Nowhere to Fangletrap to 10 east of Snonxburgh to 20 northwest of Bigwich, an east-west line of thunderstorms 10 miles wide, tops to infinity, moving from the southwest faster than your Cherokee flies" descriptions — too much information to process and at least two of the corners of the box are beyond the boundaries of my paltry mental geographical database. With datalink, just punch in the request and quicker than you can say "Center, I'd like to leave the frequency for weather information," you'll have a graphic on exactly where the bad stuff is and whether you'll be able to circumnavigate.

While this demonstration was limited to precipitation pictures, in the near future it may be possible to display areas of marginal VFR and IFR, icing conditions, potential turbulence — virtually anything that can be displayed as a graphic. Pilots could see anything that FSS or an ATC controller has, in almost real time and without any voice contact.

Last year there were 88 weather-related accidents. These should perhaps be called judgment accidents. With tools such as GWS becoming available in a few years, we just might be able to convince a significant number of pilots that the weather really is not going to get better and that landing is the only sensible option.

The FAA, MIT, and their partners in this demonstration — ARNAV, Bendix/King, and WSI — are to be commended for putting together a practical project that is designed to help light aircraft pilots. The Air Safety Foundation was pleased to test it in the real world with real pilots. The next step is to bring a commercial project to market.

The economics of using an MFD that functions as a moving map for GPS data while displaying graphical weather and traffic information looks potentially quite cost-effective. Datalink is one of those high-tech applications that, like GPS, is eminently retrofitable to light aircraft at a reasonable price, and it will enhance our decision-making capabilities significantly.


See also the index of "Safety Pilot" articles, organized by subject. Bruce Landsberg is executive director of the AOPA Air Safety Foundation.