The title of this post comes from the opening of the 1966 Star Trek TV show. Star Trek chronicled the adventures of the star ship Enterprise as it voyaged on its 5 year mission of exploration. It was kind of like the Beagle, a British naval ship that undertook an actual 5 year mission of exploration with Charles Darwin aboard. Darwin was the resident Botanist, akin to Spock, the Science Officer. Darwin restricted himself to scientific activities, whereas Spock did little or no Science. I would describe Spock's activities as more of a spy, or even more accurately as a CIA station chief, someone who coordinates intelligence gathering activities. Spock was also more actively engaged in the fighting than an actual scientist would be, certainly more actively than Darwin ever thought of being. Another difference is that the Enterprise's mission ended up lasting only three years before the show got cancelled.
And three years is just about how long the golden age of the "manned" component of the scientific exploration of space ended up lasting. The Apollo moon landings yielded a bonanza of scientific results. All of the astronauts who walked on the moon received training in Geology. Harrison Schmitt (Apollo 17 - the last moon walk) was an actual geologist, having gotten his PhD from Harvard. This geological training allowed the astronauts to select interesting samples for return and to properly document the samples and other geological items of interest that they encountered in the local area around the landings. This "golden age of manned scientific exploration" lasted from July 21, 1969 to December 14, 1972, just over three years. Other than this, the manned space program has been pretty much a bust from a science perspective. A lot was learned about how to build and operate large rockets but this could have been learned from an unmanned program. And a lot was learned about how to get humans to and from space and keep them alive while they are in space. But this is more "means" information (how to do the experiment) rather than "ends" information (what the results of the experiment are and what these results mean).
The Star Trek (fictional) and Apollo (real) programs are how everyone always expected that space exploration would take place. You would send daring men and women off to explore. That was how it had always been done. It didn't occur to anyone that there might be another way. And until the advent of Star Trek and Apollo (the '60s) there really wasn't an alternative. The word "robot" was coined in 1920. The idea of robot-like creatures is older but for a long time no one thought seriously that a practical robot could actually be built. The early post WWII era saw the emergence of the concept of "cybernetics", the use of devices to exert autonomous complex control. By the '60s these techniques had evolved to the point that unmanned "robot" spacecraft were used as pathfinders to set up for later manned missions.
The Russian "Luna" program succeeded in 1959 in crashing a man made device into the Moon. The U.S. successfully crashed Ranger 4 into the moon in 1962 and soft landed Surveyor 1 in 1966. There were also various "fly by" and "orbiter" robotic missions to all kinds of places during the same time period. Besides paving the way for a successful manned landing these robotic missions gathered a tremendous amount of scientific data. In fact, by the 1969 Apollo 11 manned moon landing, robot space craft had become quite adept at pulling off scientific missions. But the mind set was still that these robotic missions were pathfinders, paving the way for later "serious" exploration by manned expeditions. Why?
People are creative and flexible. They can make "on the spot" decisions. It seems manifestly obvious that a manned expedition should be able to do a better job. Human hands connected to human eyes are very sophisticated tools. But ultimately the argument comes down to flexibility. People are just so darned flexible and things never go exactly as planned. So having people on the scene should allow last minute adjustments to be made resulting in a more successful expedition. And until the advent of small solid state computer based technology there was a lot to be said for this idea. People will always be more flexible than robots but the "value added" has gotten much smaller. It's still there. I have no doubt that all things being equal, people will do a better job than robots. But all things are not equal.
Putting people in space is fantastically hard and expensive. People are completely unsuited for space so you have to cocoon them. They need an earth-like environment with clean air and a ready supply of food and water. You also have to deal with their waste products: gaseous, liquid, and solid. They also can't be banged around too much nor exposed to too much heat or cold, nor exposed to too much radiation. They are big (5' is unusually short) and heavy (100 lbs is pretty light). Hauling the people, their environment, their "consumables", the equipment necessary to maintain their environment and deal with their waste products, all together this results in the need to haul around a lot of tons of stuff. Getting a ton of stuff to space costs roughly $20 million dollars (e.g. $10,000/lb). And you need from a few tons to a lot of tons of stuff. And I haven't even figured in the cost of making the stuff you need to haul around.
Robots based on computer technology are much less demanding. Robots work fine in a vacuum so you don't need an atmosphere. They don't need fuel in the traditional sense, just a little electricity. They can stand a much wider range of heat and cold so you don't need nearly as much temperature regulation machinery. They need to be protected from radiation but here too their tolerance range is much broader than humans. The result is that you can build something that will do something useful that will be less than a cubic foot in volume and will weigh as little as a few pounds. The only "consumable" it will require is a little electricity and, since you can put the whole thing into hibernation for months, perhaps years, mission duration can be measured in decades. You give up some flexibility but look how much you get back. And do you really give up flexibility?
The common wisdom is that robot missions are inflexible while manned missions are flexible. I do not believe the record supports the common wisdom. If nothing else, manned missions are severely limited in their duration. This means that even if you would like to extend a mission, or even significantly reprogram it, you can't. All that stuff that is necessary to keep people alive doesn't let you make big changes in the mission. Finally, there's that ultimate inflexibility. All your astronauts have to make it back and make it back alive.
What I am now going to do is look at robot missions that are notable for their flexibility, missions that were drastically reconfigured from their original mission plan after they were launched. And I am going to start with the mission that figured in the first Star Trek movie in 1979, the Voyager mission. When launched in 1977 Voyager 1 and Voyager 2 were just supposed to visit Jupiter and Saturn. After launch the missions were reconfigured. Voyager 1 was reprogrammed to also fly by Titan, a moon of Saturn. Voyager 2 was reprogrammed to fly by Uranus and Neptune after completing its flyby of Jupiter and Saturn. Finally, both Voyagers were reprogrammed to research the edge of the solar system. This final reconfiguration was made over 12 years after the missions were initially launched. In 1990 Voyager 1, which was 3.7 billion miles from earth at the time, took a series of "family portrait" pictures of the Sun, Earth, and other planets. From this distance the Sun looks like just another star of unspectacular brightness. And the Earth is so small and dim that you can't make it out in the picture unless you know exactly where to look. The Voyagers are continuing to collect scientific data but the "family portrait" pictures were the last taken by either spacecraft.
The NEAR (later NEAR Shoemaker) mission was designed to investigate the Asteroid Eros. The initial plan was for the probe to go into orbit around the Asteroid. Photographs would be taken and various other readings (e.g. radar, magnetic field measurements, etc.) would also be made. The mission was launched in 1996 and went into orbit four years later. The package weighed 800KG at launch and just under 500KG when it went into orbit. So it consumed only 300KG (about 700lbs) in four years. Just try keeping a human alive for 4 years on 700lbs of anything. The power supply for the mission could generate 1800 watts, less than would be consumed by 20 100 watt light bulbs. The instrument package weighed 56KG, about the same as an astronaut, and consumed 81 watts of power (less than 1 100 watt light bulb). Try to keep an astronaut from freezing to death in deep space by letting him snuggle up to a single light bulb. The mission spent about a year in orbit around Eros. During this period many changes were made to its orbit to get a better look at some feature or another. But that's not why I picked out this mission.
After the probe had entered orbit and after a lot of science data had been collected a drastic change was made to the mission. A decision was made and implemented to land (well technically crash but the impact was so soft that none of the instruments were damaged) the probe on the Asteroid. The spacecraft operated for 16 days after the crash before being shut down. Bear in mind that the probe was never designed to land. All the design work was based on the probe NOT running into anything, let alone crash landing into an asteroid. But it was landed successfully and operated properly for more than two weeks after the landing.
The Deep Impact probe was launched on January 12, 2005. The plan was to rendezvous with Comet Tempel 1 (formally 9P/Tempel). An "impactor", essentially a large chunk of copper, would be crashed into the comet. The ejecta would be studied to learn more about the internal structure of comets. The plan was implemented successfully with the impactor hitting on July 4, 2005. So we're all done here, right? We did everything we were supposed to so that's it. Well, not exactly. The probe had some fuel left and was working fine, except that it no longer had a chunk of copper that could be crashed into something else. But instead of shutting everything down and forgetting about the probe the mission was reprogrammed to do a flyby of comet Hartley 2 (formally 103P/Hartley). The Hartley flyby was completed on November 4, 2010. And, oh by the way, since they had this really cool camera that wasn't doing anything most of the time, they tacked on another mission called EPOCh. EPOCh took pictures of the solar systems of seven stars that were known to have planets orbiting them and, just for fun, took a bunch of pictures of the Earth and the Moon. The probe is still going strong. There is no word yet on whether the probe will be given still more missions.
I have skipped over many examples of robot probes executing complex missions roughly according to the original plan. I have skipped over examples of things going horribly wrong and robot probes being successfully reprogrammed to recover and fully complete the original mission. These things have happened many times over. And it is possible that some missions have been lost that could have been saved had an Astronaut been on board to fix whatever went wrong. But configuring the mission in the first place to include an Astronaut would have made the mission impossible (all of the missions I have discussed above were "one way" missions, for instance) or made the initial cost of the mission prohibitive. Just sending one set of Astronauts to the International Space Station, let alone billions of miles beyond, costs more than the cost of a number of of the individual missions I have discussed.
What I have hoped to demonstrate is that taking the man out of the can does NOT mean taking the flexibility out of the mission. Because although we have taken the man out of the can we have not taken the man out of the mission. All of these missions have been directed by people from the ground. These people are not in a position to lay hands on the hardware. But they are in a position to do a lot of things. What is given up besides the ability to lay hands on the equipment is the ability to react quickly. But what is missing from manned missions is the ability to react slowly. With robot missions the people on the ground can spend days, weeks, sometimes months figuring out what can be done and how to do it. They also have ready access to colleagues, libraries, experts, etc. who can pitch in to help.
Robot missions give you the ability to do long duration missions. Voyager is still going decades after it was launched. Missions have been reconfigured years after they were launched. The warranty for the mars Spirit and Opportunity rovers was 90 days. The "edge of the envelope" estimate was that they might keep working for 270 days. The Spirit rover may no longer be working but it was known to be working on March 22, 2010, or 2210 days after landing. This represents a lifetime that is almost 10 times the "edge of the envelope" longest forecast lifetime. As of December 15, 2010 (day 2450) the Opportunity rover is still going strong. There is every reason to believe it will break the "10X" barrier. These kinds of things are just not possible with manned missions.
It is also important to understand that the capabilities of the "robot" part of robot probes are improving exponentially. The "computer" in each of the Viking probes had less capability than is now typically found in watches. By the time Spirit and Opportunity were built (circa 2003) the capabilities of the on board computers had improved tremendously. In early 2007 a major software update was downloaded remotely. This update made the rovers far more autonomous. They now make a lot of routine decisions locally without having to check in, and therefore without the delay. And the hardware used in space probes follows the "ground based" market but lags it by several years due to reliability and other concerns. So by looking at what's currently available here on the ground we know what the on board capacity of the computers on future probes will look like in a few years.
Today "quad core" processors running at 3 GHz are available inexpensively. Loading 64GB of RAM into a "server" system is normal and Flash Drives holding up to 256GB are available for less than $1000. These "available now on a desktop near you" capabilities represent what we will see in probes launched a few years from now. And it will all fit into a few cubic inches of space and use little power, almost no power when hibernating. This is enough computing power to enable some serious local decision making. Lack of local computing power will continue to be less and less of a constraint on future probes. The biggest constraint on outer solar system probes will not be computer power but the fact that you need a large dish to keep the power requirements of the on board transmitter reasonable.
The match up between manned missions and robot missions already heavily favors robot missions. And the balance will continue to tilt even more steeply in favor of robot missions. In spite of this the case for robot probes is rarely made. Some wise old man, frequently a retired astronaut, shows up regularly in print or on TV to extol the virtues of the manned space program. I can't remember any print or broadcast outlet giving time to someone to advocate for more robot exploration and less manned exploration. And that's too bad.
No comments:
Post a Comment