One Small Step for People: A Space for Humankind

Lee Dembart is a Times editorial writer.

Even before Alan B. Shepard Jr. came back from his first spaceflight in 1961, some people were wondering whether it was necessary to send astronauts into orbit at all.

Throughout the 1960s, when putting a man on the moon was a major national goal, the question was largely buried or ignored. In the 1970s and '80s, however, as scientists saw their space budgets eaten up by the shuttle program, they again raised the question: Why send humans into space when smart machines can accomplish many of the same tasks at a fraction of the cost?

One reason is political. The manned space program is one of the most popular activities of government, and it enjoys wide support in Congress. Manned flights aboard the shuttle are extensively covered in the press and on television; unmanned launches go virtually unnoticed. The message to the National Aeronautics and Space Administration is clear. If the space agency wants to retain its $7-billion budget, it needs the manned space program, which is now heading toward the creation of a permanent space station to be orbited in the 1990s.

Now the space agency, along with McDonnell Douglas Astronautics Co., has completed a study of men vs. machines in space, and concluded that there is a vital, irreducible role for each. The agency says it can and will use machines and computers for everything they can do, as it has since the beginning of the space program, but there are some tasks for which people are indispensable. Artificial intelligence notwithstanding, machines are not smart enough to replace people and will not be smart enough for the imaginable future.

"Humans are better than machines when unexpected events occur," said Georg von Tiesenhausen of NASA's Marshall Space Flight Center in Huntsville, Ala., a participating member of the Human Role in Space study. Humans can improvise. Machines cannot.

"If you have tasks that can be described from A to Z step by step, then the machines are better," Von Tiesenhausen said. "They're quicker. They have more memory. They don't forget. And they're tireless. But humans can take in a situation quickly, process it and act accordingly. Humans process a number of events in parallel. If you look at a landscape, you parallel process hundreds of thousands of images simultaneously. A computer cannot do that. A computer scans one picture at a time."

In the history of America's manned space program, the machines and the humans have acted symbiotically. Some missions have been saved by humans, such as the Apollo 13 flight in 1970, aborted when an oxygen tank ruptured on the way to the moon. Other missions have been saved by the computers, such as the space-shuttle launch last June, when the machines, acting faster than any human could have, sensed a malfunctioning valve and shut down the engines before takeoff.

If the same task must be repeated over and over, or if a number of tasks have to be performed in rapid succession--such as in monitoring a rocket during launch--the machines take win, place and show. But if the task involves recognizing a pattern or detecting an unplanned stimulus or making decisions in the absence of complete information, there is no machine that can do better than a human. People have judgment. Machines do not, and no one has a clue how to give it to them.

In its study, the space agency examined 37 tasks that must be performed in space and concluded that humans are "beneficial" or "essential" for most of them.

The trouble with sending humans into space is that it costs so much. Besides providing life support systems for astronauts, the safety requirements are much more stringent than for machines. Many redundancies are built into the shuttle that are unnecessary on unmanned rockets. The NASA study projects that the cost of operating the space station will be $32,522 an hour or $542 a minute. Some scientists argue that if that kind of money were spent on the development of smarter machines, they would become a reality, with substantial applications and benefits throughout the economy.

Not so, say the NASA experts. It is not obvious that machines are cheaper. "First you have to develop the machines," Hall said.

As currently envisioned, most of the housekeeping aboard the space station will be automated, freeing the astronauts to do experiments, assemble structures and launch vehicles. If machines could not keep the space station going, the astronauts would have time for little else. "We are very much aware of the role of automation, and we try very hard to make use of it," Von Tiesenhausen said. "We attempt to use humans only in the most important areas and automate the rest of the system."

Automation has enabled the space program to progress in 25 years from struggling to loft basketball-sized satellites to routinely sending airliners into space and back. Humans alone could not have made the jump. But machines without humans would not be repairing satellites in orbit, recovering satellites and returning them to Earth, running full-scale scientific laboratories and preparing the build structures in space.

"A certain human element will always be necessary," Von Tiesenhausen said.

Copyright © 2019, Los Angeles Times
EDITION: California | U.S. & World