Jeffrey A. Hoffman, “Exploring Space with Humans and Robots” - MA Space Grant Consortium Public Lecture at MIT (5/1/2002)

[MUSIC PLAYING]

YOUNG: My name is Larry Young. I'm the director of the Massachusetts Space Grant Consortium. And one of the great pleasures that's associated with this position is the opportunity to recruit and introduce an outstanding lecturer for the annual Space Grant lecture. As I was thinking this morning about appropriate things to say to introduce Jeff Hoffman, and at the same time trying to eat my breakfast and read the paper, I saw today's copy of The New York Times with another set of absolutely mind-blowing images taken from the Hubble Space Telescope.

And it as many of you are aware, one of the people who gets prime credit for the repair of the space telescope is our speaker today, Jeff Hoffman. Jeff has returned to MIT after a long, long trip, including five voyages into space as a science astronaut and mission specialist with the National Aeronautics and Space Administration. Prior to that time, Jeff was here at the Center for Space Research as a radio astronomer.

AUDIENCE: X-ray.

YOUNG: X-ray astronomer. Well, they're all non-optical. As an engineering astronomer, he received his PhD at Harvard, was an undergraduate at Amherst College, which also gave him an honorary doctor's degree. And he is now back here at MIT working in the aeronautics and astronautics department with faculty and students. He will be speaking to us today on the subject of 21st-century space exploration with humans and robots, the 13th annual Space Grant lecture. Jeff.

[APPLAUSE]

HOFFMAN: Thanks, Larry. Yeah. NASA's budgets are tight. And the fact that I'm here means that we don't have any travel expenses for this Space Grant lecture. So that's a big step in the right direction. I noticed for the department lunch, they served some petit fours. Now, before I came here, after I had left the astronaut office, I actually spent four years in France representing NASA in Europe, working at the US embassy. And it was nice to see the petit fours for lunch. But then I realized that here it is May 1, and we're all here working. So there's no question that we're not in France anymore.

But this is actually my first big lecture here at MIT to the department and other people. So really, I'm delighted to be able to share some of my views on exploring space, which is what I really consider-- I've devoted a major part of my life's work to. And since I'm going to discuss the role of humans and robots in the exploration of space, I should explain at the beginning that I consider exploration to be everything that expands the realm of human experience and of human consciousness. And despite having devoted a major part of my career to human spaceflight, I want to state at the beginning my belief that the vast majority of space exploration has been and always will be performed by machines.

And so it's important to appreciate the uniqueness of our biochemical based minds but at the same time, to understand how our minds interact with machines to project our consciousness to regions that we can't physically explore because exploration doesn't have to be physical. I started my professional career, as Larry said, as an astronomer. And I've always considered astronomy to be exploration, even though astronomy by definition is the study of wherever it is that you are not.

And while astronomy may be passive compared to actual physical exploration, I like to remind people that while Columbus, a great explorer, gave us a new continent, Galileo actually gave us whole new worlds. And Galileo, as you may know, never set foot outside Italy. And in our lifetime as human beings, we've explored the surfaces of all these worlds that Galileo discovered and many others besides, and we've never physically traveled any farther than our own moon.

So if our telescopes, in some sense, are the metaphorical vessels which carry our minds out to the stars, then it's certainly fair to say that Hubble is the flagship of the fleet because it has taken us deep into space and back in time to the youth of our universe. We've all seen this famous deep field photograph, which was taken by having Hubble point into a tiny part of the sky, which as far as other telescopes were concerned, was basically empty. And of course, it revealed a multitude of unseen galaxies very far away. And I show it as a reminder that, if this is not exploration, then I certainly don't know what is.

And of course, Hubble has many partners up in space exploring at other wavelengths, as well as very powerful partners on the ground, which are using some incredible techniques to penetrate the atmosphere, which I think would seem almost like magic to Galileo, or even to Hubble. Now, an important theme of my talk is going to be that exploration goes beyond just scientific research into other realms of human experience. And this is certainly true for astronomy, which engages not only our scientific curiosity but other aspects of our humanity-- our intellects, as they travel far away with the ever-increasing gaze of our telescope.

But I think they bring our emotions and our spirits with us. And that's why we get on the front page of The New York Times pictures of the Hubble telescope. That's why the public has responded with such enthusiasm to these magnificent pictures, like this famous picture, which I look at it as a picture almost of creation itself. It's the biblical, let there be light, if you will.

Well, I've begun my talk briefly by talking about astronomy because that's where I began my own interest in space in my career. But I don't propose to dwell on this aspect of space exploration, except to mention my own very good fortune in having played a role in repairing the initially flawed Hubble telescope. And this will be a jumping off point for some of my observations about humans and robotics.

Having been both an astronomer and an astronaut, for me one of the most satisfying aspects of working on Hubble was being able to unite the world of space astronomy, which generally prefers automated spacecraft, with the world of human spaceflight, which I've come to know quite well through my own personal experience. And despite my statement at the beginning about the importance of machines in exploring space, I do feel strongly about the value of human presence. And I intend to talk about the human experience of spaceflight and the human dimension of space exploration but not in the context of man versus machine or humans versus robots, which I really consider to be a false conflict.

Robots, like all other machines, are tools. We use them to expand our physical capabilities when we work with them. And we use them to operate remotely in places that we cannot go ourselves. And the relations of humans and machines has certainly evolved as our machines have become more sophisticated. And this evolution will undoubtedly continue with great benefits for the exploration of space, in which telescopes and satellites, probes, robots, and humans all have roles to play.

Now, when people first find out that I'm an astronaut, almost inevitably, the question that they ask me before any others is, what's it like in space? It's a very simple question, but think about it for a minute. What's it like in space? This is a very human question that we would never think of asking a machine, even though our automated space probes do a superb job of measuring temperatures and pressures and many other aspects of their environment.

And this, though, is not what people are asking. They want to know, what is it like to be in space? Our probes send back data, but humans can share experience. A very important part of the tradition of exploration is to report on your travels, your discoveries, to share your experiences with other people, and ideally, to make the new territory that you explore part of human culture and human consciousness.

So before I talk about robotics, let me honor this tradition of exploration and share a little bit some of the human experiences of this new environment of space. My first spaceflight, which was in April of 1985, made me the 162nd human space traveler, of which the 400th blasted off just a little over a year ago. Now, this is a large enough number of people to make going into space perhaps no longer the cutting edge exploration that it once was.

But nevertheless, anybody who who's sitting on top of a loaded rocket ready to make their first trip into space can very rightly feel that they are embarking on a personal journey of discovery and exploration. And after all, you can explore a new environment for yourself, even if you are not the first human being to go there.

So what do you see and feel and do when you go into space? Well, first of all, it is an absolutely incredible ride. Now, I hope that someday we're going to have gentler and safer ways of getting into space-- cheaper as well. In a sense, I'd like to think that two of my flights, which were devoted to testing the tethered satellite system, combined with advances in carbon nanotube technology and certainly some imaginative engineering, will ultimately lead to a space elevator, such as Arthur C. Clarke described in his book The Fountains of Paradise, which I strongly recommend if you're interested in tethers and elevators in space.

I mean, wouldn't it be great to go into a cabin and press the 42k button, and then a few hours later, you step out and you're at geostationary orbit? Certainly, the long term exploration and utilization of space absolutely depends on making access safer and cheaper. But for now, if you like excitement, well, it is a pretty great adrenaline rush when the boosters light and the ground falls straight away from you, and you break the sound barrier in 45 seconds going straight up. And then just in another few minutes, the big blue sky of Earth has turned into the blackness of space, and there you are.

And you look out the window, and the first thing that you notice is that your whole perspective has changed. You fly on an airplane and you look out, and you can see entire cities below you. But from an orbiting spacecraft, you can see countries or even whole continents, like the Himalayas and Tibet here to the right, India to the left. I'll never forget during our Hubble mission, where we were up at 600 kilometers, which is about as high as the shuttle ever goes, and we were flying right over Houston in the middle of the night. And I could look out one window and see the lights of Los Angeles and the other window at the same time and see the lights of the East Coast, with Washington, New York, and Boston.

Now, during that Hubble mission, we were pretty busy. But every once in a while, we had to wait a few minutes while we were outside while the ground did some checks on a piece of instrumentation that we had just installed. And during one of those moments, the arm operator flew me out over the wing so that I could take a photograph of what the entire payload bay looked like, so I could share with people an astronaut's eye view of what it's like being in space.

Now, this is not an advertisement for the Flat Earth Society. I can tell you from personal experience the Earth is round. I circle it every 90 minutes. But of course, wide angle lenses produce some interesting optical effects, as does, actually, the highly curved reflective surface of the telescope itself. And my favorite of those images is this one, which I call the transparent Hubble telescope. So you can see the Earth right through the telescope. And I will leave the explanation as an exercise for the students.

But for everybody, as you look at these pictures, quite seriously, trying to think about, how would you program a robot to look for humor in the environment that it's exploring, or beauty, for that matter? And talking of beauty, when we installed new magnetometers, we had to go all the way up to the top of the telescope, which was about 15 meters-- I don't know whether to say above the shuttle because you can see the Earth down below us. But in any case, we were moving pretty far away from our home territory.

And for me, this was one of the most spectacular and emotional moments that I experienced in space because there I was-- I'm the one with the red stripes around my legs. I was floating between heaven and earth, really, attached to the arm just by a slack stainless steel cable. And every few minutes-- again, when we had a little bit of spare time while they were doing the checkout on the new magnetometers-- I could let go and become a free floating satellite. And it was really quite an extraordinary feeling, especially if I turned my back to the shuttle, in which case I really had the feeling of kind of being lost in space, which was quite exhilarating, but certainly, no place to be for anybody who suffers from vertigo.

And I'll remind you that crews who are going to be spacewalking outside the International Space Shuttle are going to see the shuttle even farther away than we did, as the station gets bigger and bigger. So that's a little bit about what it looks like in space. What about the question of, how does it feel in space, which is obviously more subjective and harder to communicate. Well, I'm sure you can appreciate the initial euphoria that I felt as soon as the engines shut down after my first launch because after all those years of dreaming and waiting, I was actually in orbit and weightless in space. I actually couldn't stop smiling for about 15 minutes.

I looked out the window to see how fast we were going. But of course, there's no wind. There's no noise. There's no vibration. And so there's no real feeling of speed. But nevertheless, when you fly over San Francisco and you look at your watch, and 10 minutes later you're over Cape Cod, you know you're going fast.

Now, there are some very strong physical sensations involved in the initial experience of weightlessness, not all of them pleasant. Right when the engines shut off, I felt like I was hanging upside down. Now, of course, there is no up or down in space. But that's the sensation you get because without gravity pulling blood down into your legs, it does tend to rush up into your head.

I'm sure everybody here has probably hung upside down for a little while. But have you tried doing it for hours or days? You can get quite a headache, really. And also, when you release gravity's pressure on your spinal column, it actually increases its height by an inch or two, which may be very ego boosting if you're just a tad under six feet, but it probably has a physical effect similar to what happens when they put people on medieval torture racks and tried to stretch them.

And then, of course, there's space sickness, which results from a total confusion of your body's vestibular system, which is responsible for maintaining your balance here on earth and suffers from the lack of the expected gravity induced cues when you get up into orbit. Now, actually, we can treat most of these unpleasant symptoms pharmacologically. But you should realize that the initial accommodation to weightlessness does involve significant stress.

But humans are remarkably adaptable. And after a few days in space, just about everybody gets over the initial problems of adapting to this new environment and starts really enjoying the experience of actually being in space-- floating, weightless, looking at the world beneath you, space all around you. And everybody, of course, reacts in their own way. But I found that being in orbit almost universally evokes a sense of awe, real awe, in the deepest sense of the word.

I love climbing mountains. And for me, I think it's not dissimilar to what a lot of people feel when they go into the high mountains, where you have to work hard in an unforgiving environment and take certain risks to gain the summit. And on the high ground, your view is expanded. Your mind is kind of cleared.

But of course, space is higher than the highest mountaintop. And the environment is even more unforgiving. And the most powerful and unique feeling in spaceflight you cannot get on a mountaintop, and that's the physical freedom of weightlessness itself, which is an utter delight. For me, the weightless experience goes beyond being just physical. It has psychological, emotional, and maybe even spiritual dimensions.

And unfortunately, we don't really possess the language to adequately express this because we don't have the shared experience. And I find that even pictures can't really convey the inner feeling of weightlessness. So I try to use pictures symbolically.

Have you ever dreamed of flying? A lot of people do. But this magic carpet is not an illusion. We do it. Weightlessness gives you this wonderful sense of freedom. It's an ability to do outrageous things which you can barely imagine on the Earth.

And this joy of weightlessness is one of the main reasons that I believe that humanity really has a future in space because I'm convinced that large numbers of people are going to want to explore space personally. And they'll be willing to pay to do it, and they won't be content just to send robots. And actually, the development of the infrastructure necessary to transport and support large numbers of people in space will end up being a huge boon for space exploration beyond Earth orbit.

Now, I have to make a disclaimer. I still work for NASA. I have to remind you that NASA does not send me or my colleagues up into space so that we can see beautiful views or experience the physical bliss of weightlessness. There's no question. I'll carry moments like this in my memory for all my life, but we were sent up there to do important work on all of our flights. And I'm glad to say that we did.

Nowadays the work that most astronauts are engaged in when they go into space involves the International Space Station. Now, my talk formally is about space exploration. And I have to say that what we're doing in low Earth orbit these days is, in a sense, not so much about exploring as learning how to live in space comfortably and take advantage of this unique environment, while at the same time trying to figure out eventually how to explore further away from the Earth. But since the space station is human spaceflight, at least for the foreseeable future, I should make some comments about it.

First of all, scientific experimentation is one of the primary functions of the space station. And most of us have worked in laboratories on the Earth here, and we know what it's like to conceive and develop and carry out new experiments. When we find ourselves faced with repetitive tasks, we generally try to automate the process because nobody likes to waste time. But we don't automate laboratory experiments until we thoroughly understand the systems and we've worked out all the bugs because it's almost always the case that during the initial phases of getting a new experiment working, there are surprises. And when you have to do a lot of troubleshooting, you need to be personally involved with your equipment.

The time for automation is when you understand the hardware and the software and the processes. And that's how terrestrial laboratories work. We all understand that. And I have to say that it amazes me when I hear people saying that laboratory type research on the space station could just as easily be carried out automatically as by having humans involved. Now, as I said, scientific research is an important goal of the space station. But it's certainly not the only one.

And the station is going to be used extensively as a testbed for new space technologies, I hope, such as being able to try out more efficient solar power generation systems, for example, or deployment mechanisms for large foldable antennas, or closer to home, control algorithms for large multi-satellite constellations before we have to commit to using these technologies for new and expensive autonomous systems. But of all the things that I hope the new space station can become, the most important is as a testbed for new space robotic technology because this is going to be one big space station.

Now, we're going to work hard to figure out how to increase the number of people living and working here. But even if we had 10 people on board, when you look at all the laboratories that are imagined for here, if we really fill up all this lab space with experiments, then robotic technology is going to be absolutely essential to allow human beings to fully utilize the opportunities that the station presents. And of course, robotic technology is also going to be essential to eventually allow us to explore further away from Earth.

This is the place where we should be testing robotic servicing techniques that eventually we can use to allow repair robots to go out to the Earth Sun Lagrangian points, for instance, to service what we expect in a decade or so are going to be fleets of astronomical, solar, and Earth observing satellites, much in the same way that we astronauts serviced Hubble. Now, after we repaired the Hubble Space Telescope, I was often asked whether robots couldn't have performed these repairs as well as we did.

Well, the answer was, and still is, no. They could not. Neither then nor now do we have space robots sophisticated enough or with sufficient tactile capability to do a lot of the fine work we carried out. As an example, replacing the solar array drive electronics, we were dealing with tiny, non-captive screws on the back of the electronic connectors which kept coming loose. And they were floating all over the place like a swarm of flies, and we had to keep grabbing them with our fingertips, putting them in the trash bag and eventually chasing one of them with me on the end of the arm hanging out. It was quite exciting-- certainly, nothing that we were expecting and way beyond the capability of any robotic system right now.

However, I hope that this is not always going to be true. As I said, robots are tools. And the better our tools are, the more useful work we can do with them. Let me introduce you to Robonaut, which is a research project underway at the Johnson Space Center. And they're connected with a lot of university groups, including some people here at MIT.

The idea here was to make an anthropomorphic robot that could use the same tools that EVA astronauts use so you don't have to develop a whole new robotic tool suite. And I visited the laboratory several times. The results, I have to say, are very impressive, especially the dexterity of the hand. The guy who is the chief designer here actually gave a talk at the American Society of Hand Surgeons.

My brother is an orthopedic surgeon, specializes in hand. And he heard him, and he said it was absolutely fascinating the way that they've built similar tendons and actuators to really try to mimic the motions of the human hand. And I saw a very impressive demonstration in the lab of Robonaut actually using an EVA tether clamp to put on and off a hook. And I have to tell you, this is not an easy thing to do with an EVA glove on, and Robonaut was able to do it.

The biggest limitation currently is the lack of an effective force feedback system. And the operator told me that when he went to grasp something, he had to use primarily visual feedback, and that it was really like trying to work-- what it would probably be trying to work-- with your hands being totally numbed by a local anesthetic, where you really can't feel anything, and you have to work just by what you can see. There are force feedback systems available, but currently they're much too bulky and not nearly as sensitive as our human nervous system.

But progress is being made. And hopefully, the situation will improve. And I certainly look forward to someday seeing Robonaut get a test flight on the International Space Station because if we don't eventually have robots capable of the sort of repair work that we did on the Hubble Space Telescope, then that will mean that progress in robotic technology will have stalled out, which would be a huge pity for space exploration, because no matter how much progress we make in human spaceflight, there are always going to be places that we cannot go ourselves physically. And the better robots we have, the more sophisticated exploration we're going to be able to carry out.

Now, there's another area where human presence has an advantage, and this transcends just being able to manipulate things manually. While repairing Hubble, we had several surprises that forced us to adopt procedures that were completely different from those planned before the mission or suggested in real time by mission control. For example, these doors that you see underneath the wide field planetary camera, which contain the gyroscope compartment, we opened them the first day we went out there. And then when I went to close them, they wouldn't close. And that would have absolutely crippled the telescope if they were left open.

And so I spent about a half hour playing with the doors, trying to figure out what the problem was. And it was hard to figure out what we were going to have to do. It required a combination of looking from lots of different angles, feeling how the different latches were engaging, wiggling things, and also-- and here it gets a little bit more complicated to explain. But somehow after doing all these things, then being able to step back and get the whole gestalt out of the situation and use my years of training and experience to hope for an inspiration, which we did get, of how to adapt a standard EVA tool and use it in an unconventional manner to fix the problem.

So we described the problem in our proposed solution to mission control in words, and we sent down lots of TV pictures to give them the best-- how should you say it-- the best virtual impression of what we were seeing. But somehow the people on the ground just didn't seem to get it in the way that we did. They came up with lots of suggestions, which we didn't think would work, and one which we even thought would damage the telescope.

And they were afraid, on the other hand, that what we wanted to do was going to damage the telescope. And I'm not really sure if even the best virtual reality tools that exist could have bridged the gap. In the end, after considerable discussion, the flight director made a command decision that since NASA had gone to a lot of trouble to train us and send us up into space, we were there and could understand the situation better than they could on the ground. And in the end, they needed to trust us and let us use our judgment and get on with our plan, which we did successfully.

It's certainly true that in unplanned, time critical situations, humans are far more efficient and flexible than even our best robots. Now, deciding whether to use humans or machines for various tasks isn't limited to space. We deal with this issue in factory assembly lines and surgical operating rooms, deep in the ocean, on military battlefields. And in general, decisions about using humans or automated systems in most endeavors are made on the basis of utility and economics.

But for various historical and sociological reasons, which probably would be a good thesis topic in space policy, the questions of humans and robots in space has assumed almost the nature of a religious conflict, which I think is unfortunate. I would prefer to deal with the issue more dispassionately. And maybe looking at a few examples will help.

You take surgery, where robotic assistants already allow doctors to operate with more precision than human fingers can provide. Now, of course, these are all teleoperated systems. They should really be considered, I think, as ultra sophisticated tools for the surgeon. After all, it isn't the surgeon's finger that does the cutting in the first place. It's the scalpel. And so what do we care if there's an additional tool between the surgeon's hand and the scalpel?

And even if the operator is in a remote location, we're going to have other surgeons and assistants standing by in case something goes wrong because generally these robotic systems are designed for very specific tasks. Things can get a bit emotional if you're talking about automating a factory, and workers are afraid that they're going to lose their job. But this isn't a religious fervor. It's bread and butter economics.

And in the military field, it's certainly true there are pilots who are not happy about the possibility of losing their primary role in aerial warfare with the advent of unmanned aerial vehicles. But in all the military services-- actually, I was looking for Army, Navy, Air Force, and Marines. I couldn't find any Marine UIVs. And maybe you can't replace a Marine by a machine. I don't know.

But in any case, the decisions on adopting these automated combat systems ultimately are going to get made on the basis of cost and strategic and tactical effectiveness and of course, the added factor of trying to minimize casualties to our own troops. And finally, oceanographic research and deepwater commercial operations are another example of sharing tasks between remotely operated systems and systems where humans travel to the work sites. Now currently, ocean systems don't have a whole lot of onboard artificial intelligence. And so almost all ocean systems, whether they're tethered or not, are what we would call teleoperated.

But even the human carrying systems, such as Alvin here-- which by the way, this is a new altitude record for Alvin in the other direction. Even with these systems, you don't put human hands in direct contact with the deep ocean environment. Humans certainly have on-site presence, but they work via manipulators and other tools. I've actually spoken recently with the director of Woods Hole about this question of human and robotic choices in ocean work. And he tells me that, actually, more and more it's true, that unmanned systems are used for commercial tasks and for reconnaissance.

They're very good at taking pictures. They're cheaper to operate. They don't involve risk to humans. However, Alvin still seems to have an advantage for collecting useful rock samples and also core samples. Now, he's not sure if this advantage is going to persist 10 years or 15 years from now as robotic systems improve.

But he also mentioned another difference, which was much harder to quantify because he told me that he and his colleagues really look forward to their dives. Being inside Alvin at great depths-- and the inside, it really looks a lot like a spaceship here. And when they're in there, it focuses their minds, and it gives them a sense of the environment in which they're working in a way that they never, ever experience when they're working with remote submersibles, sitting around consoles up on the surface. And he really considers this-- he calls it ocean awareness-- that it's valuable in directing his research. And when I heard him say that, I really did feel an echo of the feelings I tried to describe to you before that I heard when I was working on Hubble.

Just before I return to space exploration, I'll end this digression into other human-robotic areas with a quote from Scientific American on deep sea exploration. And Scientific American is a pretty reasonable publication. They say, most scientists involved in exploration of the deep ocean have moved away from human occupied submersibles toward robotic craft. They're relatively inexpensive and of course, carry no risk to a human operator. Indeed, remotely operated vehicles have become so popular as tools for the offshore oil industry that economic forces could soon render conventional submersibles extinct.

But for humans to lose altogether the ability to explore the ocean depths in person would be unfortunate. Quite aside from the question of whether some subsea jobs can best be done by someone on the scene, this loss would be a blow to the human spirit of adventure. For these reasons, it seems a worthwhile goal to develop a better class of deep sea submersibles, not to replace remote or autonomous vehicles, but to offer a complimentary in situ capability for those who want it.

Well, maybe the last part of his quote sort of approaches being quasi-religious in its overtones. But I think it's basically a balanced view of the utility of human and robotic systems that could also apply to many activities in space. Personally, I loved robots in space. They have taken me as close as I'll ever get to the surface of Mars.

Like most of you, when I watched the '97 Martian Pathfinder poking its way around the surface up against rocks and finding its way against obstacles, I really had a sense of the presence of human intelligence on the surface of Mars. Because while tools were originally limited to enhancing human physical capabilities, modern technology has really made it possible to use tools to project human presence. Now, I wouldn't talk about robots competing with people in exploring space any more than I would talk about robots competing with telescopes.

Think about humanity's most distant physical explorer, which is Voyager 1, which is now about, oh, 12 billion miles away from the sun-- a little over 12 light hours, which is about 0.03 of a percent of the distance to Alpha Centauri. It's been traveling a long time, and it hasn't even left our own heliosphere. And I don't want to get bogged down in the details of the geometry of the universe, but in a sense, we can say that our robotic probes have access to about roughly one part in 10 to the 30th, maybe, of the volume of the universe that's accessible to our telescopes.

So we can dream of sending probes to the stars. But barring the realization of some science fiction space warp technology, almost the entire volume of the universe is going to have to be explored with telescopes and not with robotic probes. Now, when we come inside our own solar system, and we put telescopes and other remote sensings in orbit around celestial bodies-- oh, no.

So what do I do here? I'll just push okay. I do my best to test all these things out before we start, and then they hit me at the end. Okay, virtual memory is low. So it goes.

In any case, so here we are getting a nice look, much better than we can get with telescopes around our own Earth. And of course, they give us a much better view of the surfaces when they're in orbit. And if we have machines that can actually land and manipulate objects on the surface of another planet, this gives us tremendous power to explore, much beyond just looking or even taking in situ measurements. Good vision is essential, but touch and maneuverability takes us into another dimension of control over the environment being explored.

In places where robots can get to, they can do things that telescopes can't. And there's nothing so complicated or controversial about this. In places that people can get to, we can do things that robots can't do. It's just that it's a lot harder and riskier, and generally, a lot more expensive to send people than robots, just as it's harder, riskier, and more expensive to send robots than to look through a telescope.

There's really very little area of overlap. The farthest away from home that human beings have ever explored is the moon, which is less than a half a million kilometers away. And we didn't stay very long. And even when we get to Mars someday, which we will-- I won't say when, but we will-- but even when we get there, human beings will still have reached only a tiny fraction of the domain which is already accessible today to our robotic probes, just as our robots have only reached a fraction of the domain explored by our telescopes.

And that's what I really meant when I said at the beginning that most space exploration has been and always will be done by machines, meaning telescopes, satellites, probes, and robots. But where people have gone, we've done things that machines could not do. And we've experienced these environments as human beings.

And we can discuss the flexibilities that humans bring to carrying out scientific experiments in space. It's important. But we also need to remember that there are other types of knowledge besides scientific. And human presence can make a big difference in communicating the experiences of new environments in ways that go beyond science.

Think about the Earth's polar regions, the land of the midnight sun and of the Aurora, which we see here from space. These are areas of intense exploration and research activities. Now, I've never been to Alaska or the Yukon. Probably few of us in this room have ever been there. But I suspect that every one of us has an idea of what these arctic regions are like.

And I wonder how much of our general perception of, what's it like to be in the Arctic, comes from scientific journals. And how much comes from Jack London's famous short story, To Light a Fire? Because exploration goes beyond science. People realize this intrinsically when they're asking that question, what's it like?

To truly know an environment in human terms, and especially to communicate what that environment is like, can often require art, as well as science, and it helps to be there. Alan Bean was an astronaut who landed on the moon on Apollo 12 back in late 1969. And 10 years later, Alan was the mentor of our group of young astronaut recruits.

And I was impressed. Here's an old Navy fighter pilot. But we enjoyed talking together about literature and art. And Alan eventually left NASA to try his hand as a full-time artist. And he said he wanted some way other than words or photographs to express what it looked like on the moon, what his experiences were being there. Now, most of the lunar landscape is certainly very well represented by black and white photographs. The only color in this picture comes from the human artifacts that we put there.

When Alan looked at the rich variety of surface textures at Hadley Rille, which in this case, was explored by Apollo 15, and it was a magnificent sight to explore, probably the most visually impressive of all the regions explored during Apollo. But Alan knew what it looked like to the people who were there. And he tried to express the variety of texture through an almost impressionistic use of color.

Now, Alan would be the first person to tell you that the moon doesn't necessarily really look like this, anymore than the Rouen cathedral or the haystacks that Monet painted so many times really appeared, literally, in all those colors. But for me, Alan has showed something of what it's like to be on the moon that I never would have been able to experience just from looking at a photograph. Now, certainly, robots may not write books or paint pictures. But as I said, there will always be places that our robots can go where our bodies can't follow.

So if exploration is ultimately the expansion of human consciousness, then we need to do everything that we can do to enable our minds and our spirits to follow our robots where our physical bodies can't go. So future space exploration needs to use all the virtual reality tools that we possess to put ourselves inside our robots, to use robotic presence to extend our sensory experience, because our consciousness of the world is ultimately tied to our sensory experience of the world.

So I think, actually, the Media Lab should be playing as large a role in future exploration as places which are designing new propulsion system. When we go through the list of our senses, actually, with vision, we already can do pretty well. Given proper sensors on the lunar surface and sufficient telemetry bandwidth, we still have a little problem with peripheral vision. But nevertheless, I could stand on the Earth, look around, and have almost the same visual experience that Alan Bean had standing on the moon.

And sound isn't going to be a problem if we're in an environment where sound is important. The ill-fated Mars '98 Lander actually had a microphone on it, which would have let us hear the wind and the dust on the surface of Mars, and maybe even the sound of the shovel digging in Martian soil. And eventually, we will have a microphone there to complement the visual with the auditory sense experience.

The senses of taste and smell-- well, I don't know if any of you are familiar with Norman Mailer's book about the Apollo program. It's called Of a Fire on the Moon. It's an interesting book. In it he says that ultimately, to know the moon as human beings, we need to be able to smell a lunar rock.

I think he had the same idea in the mind about sensory experience. It's a nice thought. I'd love to try it if I could get my hands on a lunar rock. But in general, taste and smell play such a small role in human society that we've never even bothered to develop technologies to transmit these senses over a distance. So I certainly don't see the lack of virtual smell and taste as a significant barrier to space exploration.

But touch, the fifth sense, of course, is something else completely because as I already said, the ability to manipulate the environment that we're exploring is critical. So to really expand human presence into a new environment, we need to incorporate sophisticated haptic-- that is, touch feedback systems-- into our space robots. Given sufficiently sophisticated visual and haptic feedback, I can certainly imagine putting a mobile version of Robonaut on the surface of the moon and having back on the Earth a good part of the experience, not just of seeing but now, actually, of exploring the lunar surface.

Yeah, we have a two-second roundtrip light travel time, but that's short enough so with sufficient training, you could guide a robotic partner around the surface of the moon almost as if you yourself were there on the spot. And you would eventually come to feel that a significant fraction of your consciousness was present on the surface of the moon. That's the nature of virtual reality, to create a convincing, internally perceived environment. Not only would it be great fun, but it would be an incredibly powerful tool for lunar exploration.

But when we extrapolate this to Mars, not to mention farther out in the solar system, we have to face the problem of the finite speed of light. Now, I've talked with robotic specialists about predictive algorithms, optimized feed forward loops, and they say that if you've got all these things organized properly, the humans can actually operate a robotic system with up to a five- or six-second time lag with not too much loss of efficiency, as long as the environment isn't changing too fast. But with other planets, we're talking about minutes or hours, not seconds.

Now, we can certainly transmit back sufficiently sophisticated information from Mars to permit reconstructing a virtual presence after the fact. So you can use your computer mouse and fly virtually through Martian canyons. It's really exciting. But the idea of interactive virtual presence changes fundamentally as you increase the communication travel time.

Flying a computer mouse through a virtual canyon is a lot different from controlling a real airplane flying through real canyons, where you can really crash into the walls. Now, obviously, from a strictly technical robotic point of view, what we want to do is build more intelligence and autonomy into our robots so that they can do a maximum of useful work on their own while they're waiting to communicate with us. But the degree to which autonomous robots far away in the solar system are ever going to give us a real conscious awareness of new environments is going to be far less than totally interactive, real time virtual reality. And of course, the speed of exploration will be a lot slower.

Now, I don't foresee this as a long-term impediment to exploring Mars because as I said, eventually, we're going to have human beings on Mars. Because it's such a fascinating place with abundant evidence of a rich geologic history, that once we have the capability to send people there for a reasonable cost, whatever that is, and with reasonable risk, whatever that is, I have no doubt that we're going to do it. And it will be in large part because the results of our telescopic and robotic exploration of Mars will have been so exciting that people will feel that they want to get to know this planet as well as humanly possible. And that's the key-- as well as is humanly possible-- because ultimately, the best vessel to carry human consciousness into a new environment is the human body.

But when we go, we're not going to throw away our robots. Just the opposite. Human explorers on Mars are almost surely going to exercise real time control over armies of exploration robots. And they'll mostly be autonomous but susceptible, when necessary, to real time assistance to overcome problems beyond their capability to solve. We don't necessarily have to develop robots that are as good at complex decision making as human beings for them to be useful explorers, especially if the human explorer is nearby to redirect them when something unexpected happens, and they don't know how to proceed, and, of course, to fix them when they break.