The visionary scientist Freeman Dyson died early this year at age 96, but he gave a terrific talk in 2011 on four revolutions he had seen: space, computing, nuclear power, and genomics:
You may know him from the Orion project, a scheme to power rockets by throwing a series of small nuclear bombs out the back, or from the Dyson Sphere, his proposal that advanced alien civilizations could be detected by the flickering infrared light from stars that were completely surrounded by artificial worldlets, or (his major contribution) from his being a midwife to quantum electrodynamics in the late 1940s, even though Feynmann, Schwinger and Tomonaga won the Nobel for it. He had been everywhere and known everyone.
So the whole video is worth watching, but let me pick up on something a questioner asked him at the end (51:35). He asked why space technology hasn’t moved as fast as computing, with its exponential gains due to Moore’s Law. Dyson replied that unmanned space tech was improving quickly, and largely because of the processing improvements made possible by Moore’s Law. When he was working on Orion, he thought that space explorers would be like Darwin in the Galapagos, tromping around collecting samples on the moons of Saturn and writing reports. He had no idea that there would be craft like Kepler, which has found thousands of exo-planets by radioing back the tiny dips in light that happen when a planet comes between a star and us.
But is this true? Can we see Moore’s Law at work in space? Let’s pick one concrete parameter, the communication rate back from space probes. The rates vary hugely with distance, so let’s pick the most common destination for space probes, Mars, and compare them:
They range from 33 bits/sec with the very first successful probe in 1964, Mariner 4, to 6000 Kbit/sec for the Mars Reconnaissance Orbiter in 2004. That would be a doubling every 2 years, but those are outliers. If we pick more typical numbers like the 16 Kbit/sec of Mariner 9 and the 2000 Kbit/sec of Mars 2020, we get a doubling every 6.7 years. That’s not the chip version of Moore’s Law that doubles every 1.5 years, but it’s decent.
The rates have improved because of:
- More power in the transmitters: probes have gone from a few hundred watts to a couple of thousand watts due to better and bigger solar panels
- More accurate pointing of higher-gain antennas. The narrower the beam, the higher the bandwidth, but the harder it is to keep it pointed at Earth.
- Better modulation of the signals. This is where Moore’s Law really matters – more complex chips allow more complex modulation of the radio signals, permitting them to be more easily distinguished from noise.
- Better ground receivers. The Deep Space Network now operates huge radio-telescopes all around the world.
The next leap forward will be laser links. This was the plan for the Next Mars Orbiter, which would have had a 100 Mbit/sec laser link back to the Earth. It would replace the current relay satellites like MRO, Mars Express, and MAVEN, which are getting old. If the Orbiter went up in 2030, it would be a 16X speedup over MRO in 26 years, maintaining the bandwidth doubling every 6.7 years.
The 12 missions shown above are the ones I could find data rates for, and were largely the NASA missions. There have 61 missions total to Mars, of which 28 (only 46%) succeeded, and 5 (!) are en route. A single launch may carry multiple missions, E.g. Mars 2020 carries both the Perseverance rover and the Ingenuity helicopter. They’ve all been done by governmental entities. Of the 28 successful missions, NASA (US) did 21, ESA (Europe) did 3, Roscomos (Russia) did 3, and the ISRO (India) did 1.
Special mention should be made of the Emirates Mars Mission, which launched a few days ago. This was planned and run by the United Arab Emirates, with the probe built by the University of Colorado. Single universities and small countries can now send missions to Mars! This is a wonderful development.
So, yes, unmanned space really is moving forward at a good rate. What hasn’t really moved is manned space flight. Dyson talked about this too at 2:53 in the video. He and Carl Sagan were on a committee to examine the science that could be done on the International Space Station. They heard 48 proposals, and found that 46 of them could be done better by unmanned satellites, since they needed different orbits or quieter environments than the ISS could provide. The other 2 were on the effects of weightlessness on human beings, so they had to admit that, yes, they had to be done on the ISS.
But Dyson wasn’t bothered by this. He liked the Russian attitude that manned space is a matter of destiny, not science. He had been to Baikonour before a Soyuz launch, and noted how the whole town would turn out to fete the astronauts, with great parades and speeches. Treat it as a matter of adventure and pride, not science.
That’s what happened with the recent SpaceX launch of NASA astronauts Douglas Hurley and Robert Behnken to the ISS. No one cared what they would actually do on the ISS – they were just relieved that the US was back in the game. (If you do care, it’s described here, and seems to have mainly been equipment upgrades and testing).
Adventure and national pride is probably enough to get people back to the Moon, but Mars is an order of magnitude harder and more expensive. We won’t see people there for a long time, but there are already 8 active robots there, and 5 more on the way. Machines really can benefit from the amazing progress of Moore’s Law, so Martian robots are doing great.