[/caption]
When you hear about robots and space exploration, the first thing many people may think of is R2-D2 and C-3PO from Star Wars. While we may not be quite there yet, robots have become a major, even necessary, part of space missions. The many probes, landers and rovers that have been sent throughout the solar system are essentially robots, which have become more advanced over time. Then there’s the new Robonaut, a humanoid robot designed to assist astronauts with a variety of tasks in space including on the International Space Station, for example. But what is next? That was the subject of a panel discussion last Tuesday at the Von Braun Memorial Symposium in Huntsville, Alabama. The future being planned by the robotics experts involved is one of both humans and robots working together in space. The future is now…
“Can we have both robotics and human exploration of space?” was the question of the day. While there have long been advocates of both, there has also been a prevailing debate over which is better; robotic missions are less expensive and don’t put people in danger, but there are some things that only humans could do efficiently and quickly. The rovers on Mars for example, have done an amazing job of exploring the Martian surface, although human astronauts could do a lot of the same tasks faster. Also of course, people can experience the wonder and excitement of exploration in a way that machines can’t.
Instead of choosing between the two scenarios, the best idea, which I personally agree with, is to do both in tandem. That was the focus and apparent consensus of the symposium, that the best way forward is for humans and robots to work together, complimenting each others’ strengths and weaknesses. Humans might be better suited for on-site detailed exploration such as sample-taking, while robots could better handle other, more dangerous jobs.
The use of robotics has become a “pervasive technology across both military and space” according to Dr. Suzy Young of UA-Tuscaloosa’s Research Office. She also cited sources which claim that robotic intelligence could start to approach that of humans by 2040. It may still sound like science fiction, but it is quickly becoming science fact. Maybe those lovable droids from Star Wars aren’t too far off now after all.
Robert and I predicted this
Well, a panel with robotics experts would happily say that, wouldn’t they?
I don’t know what is “best”. Certainly if humans are to personally explore and colonize, we would benefit from robotics.
I would say that the most difficult questions, sometimes the most interesting ones, could sometimes benefit from human presence. If the question of “is or was there life on Mars” is not trivial, we would need paleontologists that scour the planet to get a robust answer. It took decades to assess and start inventorying ecological or even worse fossil sites on Earth.
That is not going to happen anytime soon, and I think most biologists would say the same; maybe they should contact their astrobiologist friends.
Intelligence, whatever it is and biologists are increasingly uncertain on that score, is not merely a matter of processing power and memory capacity.
If we want to emulate a biological mind to get a toe into the waters, it is embodied. Our mind is not situated in the brain, but the brain-body interactions, where the body is heavily enervated and independently hormone controlled.
In principle all wetware can be emulated in software, but note that the “all” is not an eventual target but more or less the obligatory threshold to get the system “minding”. After that bottleneck pass, which can take decades if not centuries more to unravel*, it would be a cinch to have AIs exploring the phase space of constructions of AIs and enter an exponential development regime.
But most likely not until then.
————
I am assuming it can be done, but I don’t see why not. In worst case we could start all over and take species with a rudimentary central nervous system and evolve them for body-brain intelligence to learn how it is done and what goes into such a system.
After all nature has managed that several times (some mammals, some cephalopods).
The difference is that we are now talking millenniums or more instead.
In layman’s terms, that means: God gave man a brain and a penis, but not enough blood supply to operate both at the same time.
Too bad we’re not more like the Octopus, with brain neurons in each appendage… or perhaps we are? What are YOU thinking about? LOL!
Am liking the idea that a robot might be controlled by a ‘cloud’ net (See below comment). That is, groups of computer and grey matter input making decisions and suggestions as routed through a ‘prime logic’ gate. The And/If/Either/Or/Naught decision making as decided by a ‘democratic’ electronic/human interface neural net?
Robotic exploration allows everyone to participate concurrently.
It will require standardization and collaboration. (ha! try that in today’s political climate.)
Space is no place for humans and it goes on forever that way. (so sorry to say).
There is also the corrosive tendency of warmongers to suborn technical developments and affix their parasitic tendrils to siphon funds! Talk about a headwind.
We haven’t heard much from the Robonaut aboard the ISS lately. Is this due to crew reduction or other mission/platform shortcoming(s)? Seems like with only 3 astronauts currently aboard the ISS Robonaut would have LOTS of things to do~
Well you can follow Robonaut on Twitter http://twitter.com/#!/astrorobonaut to find out – apparently the ability to tweet is not all that intelligence-dependent 🙂
Thanks Steve! I went over to ‘his’ Tweet(s) and saw the posts, now I am curious about who is ‘speaking’ for Robonaut? Or, is ‘he’ more like the new iPhone, that is, Robonaut is ‘up’ in a cloud – voice command queries synced with a computer net?
Robotic intelligence approaching human intelligence by 2040?
What is intelligence? We all know what it is like to be intelligent after the human model. We have to grow from being born, to keep ourselves alive, to develop our understanding, and, perhaps to procreate other such intelligences, and this has shaped how we thinks. It is not clear to me what a robot intelligence should be like at all. The robot may be assembled and programmed. It may be capable of very low power modes, so there is no permanent requirement for air, food, and other essentials. It could have formidable problem-solving skills, infinite patience and thoroughness, but no need for self-awareness. It would not lie, or worry about death. This would make it very suitable for exploring space, but unlikely to pass any Turing test. I would hesitate to call this ‘intelligence’ because it is so very distant from intelligence as we understand it from the only examples we currently have. What I describe is probably more like an expert system.
If robots have to work with spacemen, it my be useful to give them human-like responses. If someone shouts “What are you doing?”. The robot might might stop what they are doing and say “I am opening the airlock. I have to clean the spacecraft, and the outside may be dirty.” Computers may look at our expressions and work out when we are alarmed or angry or not sure what we are doing. But this would probably be a designed response to make the device more predictable to a human user rather than be a proper empathy.
I don’t think we have proper words for what robot intelligence will be like, let alone predict it will be working in thirty years time. I think it may be possible to design machines with intelligence, Indeed, it would be very strange if there was something that our brains could do but a Turing-complete machine could not. But we would have to be really familiar with how such machines behaved on earth before we sent them into space.