THE IDEA of flying at 30,000 feet in an airliner whose pilot has drifted off to sleep is not exactly comforting. Yet that is the specter raised by an article in this newspaper last Monday. The Federal Aviation Administration's work rules -- existing and proposed -- for airline crew members are under attack on the ground they do not sufficiently protect against the possibility that some pilots will be too tired and, maybe, too sleepy to do their jobs well.

That attack is not a new one. It has been going on for years, mostly because of the variations in the work periods of airline crew members. One pilot cited in the story, for example, began work on four days in a row at 4:15 a.m., 4:45 p.m., 3:30 p.m., and 6:45 a.m. That's a tough schedule by anybody's standards, and there is growing scientific evidence that it is far from the best schedule for high-efficiency performance. While there is no hard evidence, there is a suspicion that this kind of scheduling has produced excessive pilot fatigue, which has contributed to some airline accidents.

In recognition that the hours pilots work have something to do with safety, the FAA in 1977 began the first full review of its flight crew rules in 30 years. Early last month, it published a set of proposals that significantly increase the amount of time pilots must spend on the ground between flights and that make other changes designed to decrease the possibility of pilot fatigue. These changes are all worthwhile -- the question is whether they go far enough -- even though they will cost the trunk airlines $41 million and the other airlines (charter, commuter, etc.) $66 million in the first year. a

But the FAA did not attempt to deal with the problems of fatigue associated with changing work shifts or travel across multiple time zones. It noted that studies of how persons respond to such travel conflict, and the variation in the physiological responses of the pilots in those study groups is too great to provide evidence on which to base a work rule.

It is the refusal of the FAA to grapple with this question, which some scientists see as a major safety problem, that has brought the matter of sleeping pilots back into the news. Some recent studies of human behavior indicate that changing work schedules have a substantial negative impact on the performance of most individuals.

If that is so, and common sense suggests it is, the airlines and the FAA have a safety problem they are not facing squarely. So, for that matter, do other employers whose employees work rotating or sporadic shifts -- policemen and truck drivers, for example. Accident statistics suggest fatigued truck drivers are a greater threat to public safety than fatigued pilots.

That gives the FAA and the airlines an unusual opportunity as well as a responsibility. Since no one, including the airlines, wants pilots sleeping on the job, they must explore this matter more deeply and establish work rules -- regardless of cost -- that guarantee maximum pilot alertness. In the process; they have a chance also to establish patterns of work that could be used to improve performance in other occupations that do not function on a 9-to-5, five-days-a-week schedule.