Hearing an exchange recently of all-too-routine horror stories about mail delivery, I became curious about what the U.S. Postal Service is doing to research its way of swifter delivery of our letters and parcels.

The post office, after all, faces what the science and technology trade calls a "researchable problem" -- getting 100 billion objects a year from here to there, a job that anually costs over $14 billion and employs some 650,000 people. Surely modern research techniques could do their bit to speed up the traffic and restrain the costs. What, then, is the postal service doing about research?

The answer is that it's now doing relatively little. And next year, it's planning to do even less, though its present research and development expenditures -- a mere $31 million a year -- are by any measure among the lowest of major industries, as well as the smallest of any major federal organization. Added to that, the one and only research facility that the postal service does maintain -- located in a complex of buildings in Rockville -- not only is minuscule for a multi-billion-dollar enterprise, but also tends to be intellectually isolated from the mainstream of modern science and technology.

There's no sure yardstick for how much research is enough for getting a job done, but the postal service -- which claims to be in the communications business -- is in the sub-sub basement of research spending. In just plain dollars, the razor blade industry outspends it on research. In the carveup of the $30 billion a year that federal organizatitons now spend on R&D, the figure for the postal service comes in as a miscellaneous entry. In terms of the commonly used industrial measure of R&D spending as a percentage of sales, the postal service gets the national booby prize, with a rock-bottom two-tenths of one percent. There's no reason to expect it to be up there with the aircraft and missiles industry's 13 percent, or even the auto and truck industry's 3 percent of sales going into research. But that two-tenths of one per cent that our $14-billion mail service spends on R&D ranks far behind a number of industries that are not renowned for interest in research, such as textiles, wood products and primary metals.

While all of these, sniffing the profit that's to be had from research, have been boosting their R&D spending, the postal service is planning a $4 million cut in research next year. The inadequacy of postal spending on new ways to deliver the mails was pointed out in 1977 by the Commission on the Postal Service, which noted that the slice of revenues devoted to R&D "is far less than for private companies in the communications field." Nonetheless, the postal service is reducing its research activities, which clamoring for an opportunity to take on the immensely complex task of initiating a national electronic-mail system.

As for what the postal service gets out of the limited funds that it does spend on research, it appears that it's receiving good value from a dedicated staff that works on a variety of problems, ranging from mailbox designs to testing of high-speed sorting machines developed by outside contractors. But a visit to the postal service's research and development department also imparts the impression of a self-contained backwater in an ear when elsewhere science and tecnology are rumbling with energy and innovation.

Unlike the big laboratories run by government and industry, the R&D department doesn't go recruiting for the bright young graduates who are coveted by research organizations that are held to high standards of output. At the post office, employment is stable and promotions are usually from within. The postal service doesn't join in the widely used practice of calling in outside specialists to examine its internal research program. That's commonplace in government, the universities and industry. Both General Motors and Ford, for example, have blue-ribbon councils of distinguished scientists and engineers to adivse on their in-house research programs.

Moreover, the postal service seems to taken an extremely modest view of the role of research in a vast industrial organization. Given that it is one of the most labor-intensive businesses in the country, it's appealing to hear from the chief of R&D that his staff includes just one industrial psychologist.

My visit to the postal service's R&D labs included a symbolically telltale episode: spread out on a table were the electronic innards of a new vending machine that's being developed. An engineer dropped a coin in a slot to demonstrate the new device. On flashed a sign: Not in Service.