We see that it wasn’t until about its thirty-fifth launch to the Moon or the planets that the cumulative U.S. mission success rate got as high as 50 percent. The Russians took about 50 launches to get there. Averaging the shaky start and the better recent performance, we find that both the United States and Russia have a cumulative launch success rate of about 80 percent. But the cumulative mission success rate is still under 70 percent for the U.S. and under 60 percent for the U.S.S.R./Russia. Equivalently, lunar and planetary missions have failed on average 30 or 40 percent of the time.
Missions to other worlds were from the beginning at the cutting edge of technology. They continue to be so today. They .ire designed with redundant subsystems, and operated by dedicated and experienced engineers, but they are not perfect. The amazing thing is not that we have done so poorly, but that we leave done so well.
We don’t know whether the Mars Observer failure was due to incompetence or just statistics. But we must expect a steady background of mission failures when we explore other worlds. No human lives are risked when a robot spacecraft is lost. Even if we were able to improve this success rate significantly, it would be far too costly. It is much better to take more risks and fly more spacecraft:.
Knowing about irreducible risks, why do we these days fly only one spacecraft per mission? In 1962 Mariner 1, intended for Venus, fell into the Atlantic; the nearly identical Mariner 2 became the human species’ first successful planetary mission. Mariner 3 failed, arid its twin Mariner 4 became, in 1964, the first spacecraft to take close-up pictures of Mars. Or consider the 1971 Mariner 8/Mariner 9 dual launch mission to Mars. Mariner 8 Was to map the planet. Mariner 9 was to study the enigmatic seasonal and secular changes of surface markings. The spacecraft were otherwise identical. Mariner 8 fell into the ocean. Mariner 9 flew on to Mars arid became the first spacecraft in human history to orbit another planet. It discovered the volcanos, the laminated terrain in the polar caps, the ancient river valleys, and the aeolian nature of the surface changes. It disproved the “canals.” It mapped the planet pole to pole and revealed all the major geological features of Mars known to us today. It provided the first close-up observations of members of a whole class of small worlds (by targeting the Martian moons, Phobos and Deimos). If we had launched only Mariner 8, the endeavor would have been an unmitigated failure. With a dual launch it became a brilliant and historic success.
There were also two Vikings, two Voyagers, two Vegas, many pairs of Veneras. Why was only one Mars Observer flown? The standard answer is cost. Part of the reason it was so costly, though, is that it was planned to be launched by shuttle, which is an almost absurdly expensive booster for planetary missions—in this case too expensive for two M. O. launches. After many shuttle-connected delays and cost increases, NASA changed its mind and decided to launch Mars Observer on a Titan booster. This required an additional two-year delay and an adapter to mate the spacecraft to the new launch vehicle. If NASA had not been so intent on providing business for the increasingly uneconomic shuttle, we could have launched a couple of years earlier and maybe with two spacecraft instead of one.
But whether in single launches or in pairs, the space-faring nations have clearly decided that the time is ripe to return robot explorers to Mars. Mission designs change; new nations enter the field; old nations find they no longer have the resources. Even already funded programs cannot always be relied upon. But current plans do reveal something of the intensity of effort and the depth of dedication.
As I write this book, there are tentative plans by the United States, Russia, France, Germany, Japan, Austria, Finland, Italy, Canada, the European Space Agency, and other entities for a coordinated robotic exploration of Mars. In the seven years between 1996 and 2003, a flotilla of some twenty-five spacecraft—most of them comparatively small and cheap—are to be sent from Earth to Mars. There will be no quick flybys among them; these are all long-duration orbiter and lander missions. The United States will re-fly all of the scientific instruments that were lost on Mars Observer. The Russian spacecraft will contain particularly ambitious experiments involving some twenty nations. Communications satellites will permit experimental stations anywhere on Mars to relay their data back to Earth. Penetrators screeching down from orbit will punch into the Martian soil, transmitting data from underground. Instrumented balloons and roving laboratories will wander over the sands of Mars. Some microrobots will weigh no more than a few pounds. Landing sites are being planned and coordinated. Instruments will be cross-calibrated. Data will be freely exchanged. There is every reason to think that in the coming years Mars and its mysteries will become increasingly familiar to the inhabitants of the planet Earth.
In the command center on Earth, in a special room, you are helmeted and gloved. You turn your head to the left, and the cameras on the Mars robot rover turn to the left. You see, in very high definition and in color, what the cameras see. You take a step forward, and the rover walks forward. You reach out your arm to pick up something shiny in the soil, and the robot arm does likewise. The sands of Mars trickle through your fingers. The only difficulty with this remote reality technology is that all this must occur in tedious slow motion: The round-trip travel time of the up-link commands from Earth to Mars and the down-link data returned from Mars to Earth might take half an hour or more. But this is something we can learn to do. We can learn to contain our exploratory impatience if that’s the price of exploring Mars. The rover can be made smart enough to deal with routine contingencies. Anything more challenging, and it makes a dead stop, puts itself into a safeguard mode, and radios for a very patient human controller to take over.
Conjure up roving, smart robots, each of them a small scientific laboratory, landing in the safe but dull places and wandering to view close-up some of that profusion of Martian Wonders. Perhaps every day a robot would rove to its own horizon; each morning we would see close-up what had yesterday been only a distant eminence. The lengthening progress of a traverse route over the Martian landscape would appear oil news programs and in schoolrooms. People would speculate on what will be found. Nightly newscasts from another planet, with their revelations of new terrains and new scientific findings would make everyone on Earth a party to the adventure.
Then there’s Martian virtual reality: The data sent back from Mars, stored in a modern computer, are fed into your helmet and gloves and boots. You are walking in an empty room on Earth, but to you you are on Mars: pink skies, fields of boulders, sand dunes stretching to the horizon where an immense volcano looms; you hear the sand crunching under your boots, you turn rocks over, dig a hole, sample the thin air, turn a corner, and come face to face with… whatever new discoveries we will make on Mars—all exact copies of what’s on Mars, and all experienced from the safety of a virtual reality salon in your hometown. This is not why we explore Mars, but clearly we will need robot explorers to return the real reality before it can be reconfigured into virtual reality.
Especially with continuing investment in robotics and machine intelligence, sending humans to Mars can’t be justified by science alone. And many more people can experience the virtual Mars than could possibly be sent to the real one. We can do very well with robots. If we’re going to send people, we’ll need a better reason than science and exploration.