Search This Blog

Saturday, July 2, 2011




The famous scientist Steven Weinberg, 1979 physics Nobel Laureate, claimed that “the whole manned spaceflight program, which is so enormously expensive, has produced nothing of scientific value." This remark capped a scathing critique in which he also said that "The International Space Station is an orbital turkey…. No important science has come out of it. I could almost say no science has come out of it.”[1] And a New York Times columnist reportedly stated matter-of-factly that “Three decades after going to the moon, NASA is sending astronauts a few hundred miles above Earth to conduct high school science experiments.”[2] Nor is the value of manned exploration likely to change, for according to Weinberg, "Human beings don't serve any useful function in space," he told "They radiate heat, they're very expensive to keep alive and unlike robotic missions, they have a natural desire to come back, so that anything involving human beings is enormously expensive." This derisive view of manned exploration is widespread amongst space scientists themselves. Indeed space scientists are likely to object bitterly to projects like the space station and President Bush’s proposals to send humans to the Moon and eventually to Mars. My sympathies are with those scientists: Manned exploration has indeed been detrimental to space science. And in the short run it will continue to be so. But in the long run manned exploration will help greatly the cause of space science while providing great benefits to the human species.


Some space scientists argue that we can achieve the goals of space science better with machines than with astronauts. Their feeling is that the money and effort that could be spent on driving science and technology to explore the cosmos is in large part lost when we concentrate instead on ensuring the safety of astronauts and on developing very expensive and cumbersome life-support systems. An astronaut needs air, water, food, and protection from a hostile environment. The satisfaction of these needs requires bigger rockets to handle the far heavier payloads – as well as far more reliable spacecraft. A human being is a delicate creature. Even with the best of our technology we could not easily send astronauts into the hell of Venus or the intense radiation of Jupiter's vicinity. A trip to Mars would also be very difficult, since it would take months under constant bombardment from the solar wind, enough to destroy upwards of ten percent of the astronauts' brain cells unless the spacecraft is especially protected – in addition to the possible adverse physiological effects from such a long exposure to weightlessness. Machines, on the other hand, can go practically everywhere in the solar system, for far less money.

Proponents of manned spacecraft reply that humans can do many things that machines cannot. For example, humans can perform experiments that require great dexterity, and they can retrieve and fix satellites. That is true, but according to their critics, not relevant. First, machines can do their more limited job in places where humans cannot or should not do any job at all. This includes not only trips of very long duration and hazardous environments, but also dangerous experiments. Second, in many space science experiments the presence of man is a hindrance. For example, telescopes have to be so precisely aimed that someone moving around in the spacecraft would disturb the observations. Third, even if astronauts can retrieve and fix satellites, and build and operate industrial facilities, whereas present machines cannot, we can design our space equipment so robots or teleoperators could handle the job. This of course requires that we develop robots and teleoperators equal to the task. Thus on the whole, the impetus that technology receives is greater from exploring space with machines than from worrying about the safe transportation of astronauts. Furthermore, advances in machine operations in low orbit can be applied throughout the solar system, whereas astronauts will be unlikely to venture beyond Mars in the next fifty years, and even that looks to the critics like pie in the sky (pun intended).

The tragic destruction of two Space Shuttles, Challenger and Columbia, threw into disarray the space program and has done great harm to space science. They have clearly shown the risks both to human life and to science from too great a reliance on manned exploration. Indeed, even when the space shuttle flies normally, space science suffers, and, as we will discuss below, the space station makes matters worse. Why should we then insist on manned exploration when we can accomplish far more, and to do it far more cheaply, safely and efficiently with machines?

Exploring with Machines

Let us take a look at the two main technologies favored by the critics of manned exploration: robotics and teleoperators.


Teleoperators permit us to handle via radio and television tasks that must be carried out at a distance. For example, a television camera on a machine transmits to the ground an image of two building blocks, and a human operator makes the arms in the machine put the two blocks together by radio transmission. Interaction between ground crews and a variety of orbiting observatories has actually become routine. We can change orbits, aim cameras and telescopes, and even perform experiments. An advance in the various aspects of teleoperators -- sensors, arms, fingers, grip, and dexterity -- will increase the range of activities that we can perform by remote control in space. Teleoperators combined with robotics can go even further: The human operator would perform certain repairs, say, in a comfortable laboratory, while a robot would mimic the same actions in a far more hostile environment.[3]

But some serious problems remain. The most obvious problem is that the farther away the spacecraft is, the harder it is to run it by remote control. Imagine that a roving vehicle on the Moon comes suddenly upon a hole or an unexpected rock. Its television camera will immediately send a picture of the obstacle to an operator on Earth. But on the average it takes that signal one and a half seconds to arrive. If the human operator reacts instantly, the instruction will arrive at the Moon one-and-a-half seconds after that. Any one who has driven a car knows very well that many disasters can be crammed into three seconds, which is the minimum time that it would take for the earthbound human operator to react to the lunar environment. For a rover on Mars that time would be of the order of eight to forty minutes, and for the outer planets we have to allow hours.

There are two ways to reduce this difficulty. One is to anticipate as much as possible and build our space machines accordingly. We could, for example, provide the lunar vehicle with a computer map of the land it must travel (drawn from photographs taken by orbiting spacecraft). Any deviation from that landscape will automatically force it to stop until it receives fresh instructions from its human operator.

On the Earth, of course, an attentive human driver is able to detect a nasty pothole and get out of harm's way in less than three seconds. But to do so he uses a variety of perceptual clues that allow him to spot a hole for what it is and to tell just how far it is. The teleoperator, by contrast, is looking trough a television camera at an alien landscape: his remote vision is poorer than the terrestrial driver’s, and he does not have all the perceptual clues that his perceptual apparatus needs to come to a quick decision. The upshot of all this is that the roving vehicle must move very slowly. That is so even when the terrain is reasonably well known. When the vehicle is called upon to do some honest-to-goodness exploration, the difficulty becomes acute.

As a matter of fact, the Russians sent one such vehicle to the Moon. And NASA had plans for another at one time. But when the best hopes for its performance were so clearly surpassed by the actual performance of the astronauts, the project – called "Prospector" – died of natural causes.

The teleoperator is then at a disadvantage with respect to a human on the spot, at least for certain kinds of jobs. Not only is his camera not as good as a human eye, it does not receive the correcting feedback of the other human senses. Human beings do not just see what is out in the world and correctly report it to consciousness. They pick out and concentrate instead on a variety of subtle clues as to what is most relevant and worthy of attention. A human observer looking through a television camera has fewer of those clues (he will be missing clues that are peripheral, in the background, or correlated with hearing, smell, or touch). Even if he is highly trained, in a new situation his degraded experience may not suffice for him to recognize objects and situations in the way he has been trained to do. A laboratory biologist can tell at a glance that a guinea pig is ill because somehow its behavior differs subtly from patterns to which the biologist responds even if he cannot describe them. It is our hands-on experience that allows us to gain what appears to be an intuitive "feel" for our surroundings. Geologists and materials experts may also depend on the immediacy of contact in order to grasp the object of study in this quasi-intuitive way.

In space, it is true, many of the associations between the senses may be disturbed by the absence of gravity, and thus previous training may lead a scientist on the spot to misjudge the situation. But human beings have the capacity to adapt and to form new associations. A biologist making slides of a rat's brain can learn how to compensate for the new environment. The teleoperator, on the other hand, faces two different problems. First, at present his artificial “hand” simply does not have the dexterity to carry out that refined a task. Second, even if it did, and eventually it might, he could not use that artificial hand the way he would on Earth; he would have to be retrained so as to make that artificial hand do from Earth what the biologist does with his real hand in space.

To make a better artificial hand (or other appropriate tools) it would be wise to observe what the human biologist does in his space laboratory, and then slowly refine the technology until it is acceptable. Cutting a rat’s brain in weightlessness may be quite different from doing it on the surface of the Earth. It makes sense, then, first to try to learn what it is like to do that kind of lab work in space. That is, we need to develop a space expertise in those activities – a human expertise. Only then we might be in a position to calibrate our teleoperations.

The obvious conclusion of all this is that teleoperating technology is best developed in cooperation with human activity in the relevant areas. To perform at the level that the proponents of unmanned flight hope for, we would require to have a joint, and to some degree a prior human presence in space. This might be a useful technology to develop in the International Space Station. Although a space station is not very useful to all branches of space science, it may be very valuable to biology, materials processing and medical technology. Moreover, as our experience with the Shuttle has demonstrated, there are experiments in physics that require a high degree of finesse or complexity in the handling from space (for example, electron beam experiments to examine the interaction of charged particles with the Earth's magnetic field). Without the mission and payload specialists in the spacecraft, those experiments could not have taken place for decades, if at all.

For materials processing, one of the most promising areas of space industrialization, artistry is as crucial as scientific craft. We must not forget that the initial purpose of the space station, apart from science, is to do industrial research rather than to set up actual industries. If it were the latter, then there might be some hope for combining teleoperators and a high level of automation. But insofar as the purpose is largely one of exploration, the machine technology is not yet up to the challenge. Nor are our teleoperating abilities so developed that we could build those completely automated factories without the help of human workers in space.

[1] Kerr Than,, September 8, 2007.

[2] Reported by John Tierney, “Outer Space on Earth: NASA Should Try It,” reprinted in Detroit Free Press, August 2, 2005, p. 7A.

[3] Gilster, Paul (2002). Centauri Dreams, Copernicus Books, p. 216.

No comments: