The number of "near misses" among airplanes in the crowded skies is on the rise. Total incidents reported to the Federal Aviation Administration (FAA) -- far from all that occurred -- almost doubled in the last half of the 1970s, averaging more than one a day. By the FAA's own assessment, more than one-fourth of these close calls were "critical." As plane traffic increases in the years ahead, the problem is expected to intensify.
The FAA has been feeling the heat from Congress, newspaper editorials, and such organizations as the National Transportation Safety Board to speed up development of a workable collision avoidance system (referred to as CAS by the aviation community) and put it in place.
Yet for the moment the agency still carefully sidesteps the question of whether or not its plan on the drawing boards will ever be made mandatory.
"If we put in a system that gives out a lot of false signals, people [ controllers and pilots] will turn it off or ignore it," explains FAA spokesman Jerry Lavey. "We want to put in something reliable that will really work."
Adds Martin Pozesky, one of the FAA's top research and development specialists: "We feel if the systems are perfected, they will be implemented -- independent of any rulemaking. Everyone would want them."
Virtually everyone agrees that the present system of detecting "near misses" is somewhat haphazard. Despite the recent addition to the existing radar system of an automated device that flashes a warning if aircraft get too close, it is the already busy air traffic controller who must then issue radio instructions to the affected pilots on how to evade a hit.
In the September 1978 midair collision over San Diego, this automated "conflict alert" system was in place, but the controller thought the crew of one plane had the other plane in sight.
The search has long been on for a computer device that will automatically alert pilots flying too close to each other as to the danger and tell them what action to take without requiring the direct intervention of the air traffic controller. The current FAA program aims to achieve that goal in three phases:
* Next spring two Pacific Southwest 727s on the West Coast will begin testing an active beacon collision avoidance system (BCAS) that involves placing a sensitive instrument in the plane's cockpit that would operate independently of the existing ground radar system. Pilots of planes flying too close would be told on display terminals in the cockpit whether to climb or dive.
* Disadvantages of that system -- such as the fact that it requires a change of altitude rather than any lateral turns (pilots say they are sometimes too low to change altitudes) and cannot work in areas of heavy traffic -- would be overcome in the development of a more sophisticated "full capability" BCAS that would be ready for testing by the mid- 1980s. One added advantage is that only the affected pilots -- not all those on the same frequency -- would receive the warning and directions.
* The last phase of the FAA solution would probably not be implemented before the 1990s. Called DABS-ATARS, it links an improved transponder on the aircraft, which would supply much more information about the plane's identity and flight path, with an improved ground-air radar system.
Each of the three advances would build on the other. "None is a throwaway," says Mr. Pozesky.
Each step in the FAA approach to collision avoidance is closely linked with the ground computer network and air traffic control system that the agency manages. While most of the aviation community, including the airline industry, support that approach, the Air Line Pilots Association (ALPA) wants the FAA to take its solution one step further. The pilots who want immediate direct knowledge of any threat facing their aircraft are pushing for an additional independent airborne device as a backup.
"We want something in the plane itself in case the computer components fail," says John O'Brien, ALPA manager of engineering and operations.
He argues that such a system is technologically ready now.
The FAA did test three airborne anti-collision devices in the early 1970s but abandoned the effort because, according to Pozesky, the systems did not "test well." He argues that Step 2 of the FAA plan incorporates the best features of an independent airborne device and comes as close to one as anything else yet developed. Of continuing ALPA criticism, he says only, "They just don't want us to get too complacent."
The ALPA contends that it was opposition from the small-plane industry and internal politics that effectively quashed the FAA's search for an in-plane CAS.
The pilots' view, which some dismiss as part of the perennial battle between pilot and controllers for control of the aircraft, received a strong boost recently when disgruntled FAA employee James Pope publicly accused the FAA of "jeopardizing" safety by refusing to authorize use of an airborne CAS that had been proved effective.
Admitting that he has some sympathy for the pilot view on the need for a backup, former NTSB official and now independent aviation safety consultant Chuck Miller cautions: "The FAA does tend to believe in the adequacy of technology a lot more than it perhaps should."
Much of the controversy, notes John Galipault, president of the Aviation Safety Institute, depends on how much the aviation network is willing to trust the computer. In his view, the FAA deserves credit at the very least for an earnest effort in the search for an effective solution to the collision avoidance problem.
"I'm not saying we're going to see anything for awhile," he says, "but the FAA appears to be working diligently."