ROBOTS (Per): Welcome to the podcast, I’m here today with Stephen Gorevan of Honeybee Robotics.
Stephen Gorevan: I’m glad to be here, Per. And I’m glad you came and I’m happy to talk to you.
ROBOTS: Could you start with how you got started in robotics and your first robotics, and then we can move on from there?
Stephen Gorevan: All right. The origin of the company Honeybee Robotics was quite specific, actually. It happened one night. I was in graduate school at Columbia University, engineering department, and I had a eureka moment, literally, and it came to me, believe it or not, that I had to start a robotics company. I was reading an article about a French industrialist – this is how specific I remember this – I was in bed, and this idea just popped into my head. I can’t find anything in the article that made that happen, but it did, and this idea, [which] was perhaps the fertilizer for this situation, was that perhaps about two years earlier my partner — the president of this company, Cris Chapman, who is down the hall, who is a financial guy, and he runs our finances superbly well, and I do not know what I would do without his friendship and partnership. We went to school together, at New York University, undergraduate, and when we parted, he had to go back to Ohio, to a family business, he said “Listen”, and it was a strange thing to say, because I was just a student, I thought it was strange, he said “If you ever have an idea for a business, you tell me!” because apparently he had some kind of access to a small amount of cash, and he was, you know, I’m just a student; I have no track record. He had faith in me, if I ever had an idea. I didn’t think much about it, because I never thought I would have an idea for a small business, but then I did, that moment, and he flew out to New York, he was in Ohio, and we stayed up all night, and within days we had incorporated and he gave me a check. We started the business with two students from Columbia, and we said to ourselves the basic idea of the company, robotics, was rather new in U.S. industry, because we weren’t looking for a world-wide impact at that moment.
I had read that the manufacturers of robotics in the 80s, this was 1983 in the fall, there was a problem, because these magnificent machines were being built, but they were dumping them literally at the doorstep of U.S. industry. The industry was buying them and saying ‘We have to automate’, but there was this gap. No one on the user side knew how to use these machines properly, so we said ‘Let us be a middleman; we will take our knowledge of robotics, which was what we were studying in school, and we will be an in-between person; we will install them and we will train the user, and the manufacturers of robotics will love this because we will be selling and using their machines.’ Typically when a robot is installed in a factory, in general, it only represents about 20% of the sunk costs; there are all kinds of additional costs, machines to feed parts, conveyor belts, you name it, it just adds up. So we were able to become a systems engineering house, that put this whole thing together in a turnkey fashion for the user.
The very first time we went out to try to sell a robotic system to an American corporation, we went to a company not far from this location, mid-town Manhattan. The very first time we went out, we sold a system to American Hydron Company, which made contact lenses, out on Long Island. Driving back in a car I said “This is going to be easy.” But it wasn’t until fourteen months later that we made our second sale, so it was very, very difficult, the next year. Actually I was getting calls from people who were close to me saying, I really need to – I felt like the doctor that was trying to revive a patient with electrical paddles, and they were pulling me away saying ‘He’s gone’, that I had to leave it, but I kept on trying and trying. And, sure enough, we got someone who believed in one of our ideas, because by this time we were giving away the engineering, trying to interest customers in what we could do for them, so we were drawing up plans and being very specific. The risk was, of course, people could steal it, but I said ‘What the heck, we have got to show them what we could do.’ We found somebody that really liked what we were doing, and we got a rather large contract, that kept us in business. And we were going along fine, until 1986.
In 1986 two things happened. One, my secret childhood dream was to work for NASA. It hadn’t even entered into the business plan to work for NASA, because I didn’t think that NASA was very interested in robotics per se. Now I know that the Voyager spacecraft that had flown to the outer reaches of the solar system, these were called robotic, but to me they were really just programmable spacecraft, they were not robots in the sense that they had end-of-arm tooling, joints, that would be measured in degrees of freedom, and so on, like we typically think of robotics today. So that kind of robotics is what I was good at, and I didn’t see them in use at NASA, but I always wanted to work for NASA, and then in 1986 I for some reason became possessed with the idea of going to see someone at NASA, to see if they would…just on the off-chance; I had to try it once. Chuck Hoberman was working with one of the original workers at Honeybee Robotics, and my friend, and one of the founders, really, of the company – he is the engineer who developed the Hoberman Sphere, that expanding/contracting sphere you see in a lot of airport gift shops, it is actually an incredible mechanism. And I thought, ‘Volume is very important with respect to spacecraft, the smaller things are the better, so if you can have a deployable system that shrinks into the smallest space possible, then expands, perhaps used as antennas in space, that might unfold, or solar arrays, Chuck had some really unique ideas about how we could make really compact deployable devices, compact so they could fit in rocket fairings. And I said ‘We have to go to NASA; we really have to give this a shot.’
So we walked into the office of the head of the structures branch, at NASA Langley, Martin Mikulas, Jr., who at this time was doing the pathbreaking work on developing the Space Station, which is an incredible engineering achievement and it was also radically different from any civil engineering project that had ever happened before in human history. We stayed in his office for hours. He warned me not to come because he said he had people from Japan coming who were experts in origami, and things like that, that he had to disappoint, but we sat there three hours and explained what we could do for him, and some of the other deployable ideas we had, and lo and behold he made me the happiest person in the world; he gave us a very small but very real contract. I was so happy; I just cannot tell you how happy I was, because I was working for NASA, my childhood dream. I even have a little essay, when I was, I had to have been eight, and I wrote this little essay saying “I want to work for NASA. I want to go to the moon…” This had to be why I was so happy, and then it turns out, in that time we walked into the office, as we were walking down the hall, after we had received our first contract, Mikulas told me that “You folks happened to walk in here at the exact right time.” I said “What do you mean?” Well, what had happened the month before was the explosion of the Challenger, in 1986, and the Space Station, at this time, was on the drawing boards, but the plans to deploy the Space Station and assemble it originally involved thousands of hours of astronaut EVA activity. And, now, all of a sudden, the directive from up high was to get involved in robotics, because we cannot expose astronauts to all this EVA time, which is going to be equivalent to losing some astronauts, and right now we do not want to lose any more; we want to keep it to an absolute minimum. He has a southern accent, and he said [mimicking Mikulas’s accent] “Your company has the right name on it, because we were called Honeybee robotics.” We got a whole flurry of contracts, using real robotics, arms, end effectors, to develop a way that we could build structures on orbit, and we were building mockups, working mockups, in hangers down at NASA Langley, it was just beautiful. We had so much work that we completely stopped working for U.S. industry. We had so much NASA work, and we never looked back.
ROBOTS: Would you like to speak a bit about what you are doing in this new project that is going to land in just a week?
Stephen Gorevan: Sure. This is a little difference for us, which I am very excited about. Basically we are doing two things on the Mars Science Laboratory. First, we were supposed to have developed a super RAT [Rock Abrasion Tool], an even more capable RAT for the MSL mission, but there was a cost overrun on the whole rover development and some things had to be what we call descoped, and the requirement for a super RAT was dropped off to a brushing RAT. The Rock Abrasion Tool that we sent to Mars, one of the things that we had to equip the Rock Abrasion Tool with was a brushing system that eliminated the cuttings that we were cutting. When we cut into the rock we did not want to just have left there a pile of dirt inside the hole we ground. So, we developed a high-speed stainless steel brushing system that really brushed the rocks clean after we have ground into them. In fact you do not see any signs or any of the cuttings, in any of the targets we have gone into, these brushes have been so effective. In fact they were so effective, the brushes, that because the grinding heads were consumable and the brushes less so, although the brushes did exhibit some wear, but not much, the brushes have been deployed on their own, to brush away targets of all kinds, to just get a better look at all surfaces, and apply the instruments to where we brushed. It is sort of a poor man’s Rock Abrasion Tool. Well, these brushes were so effective that the descope that was planned for MSL, when the money got tight, we were saying ‘Okay, just make a brushing RAT.’
We have developed what is called the Dust Removal Tool, and that is now aboard the arm on the Curiosity Rover, but the second element that we have supplied for the rover is really the big one that is making me nervous like the Rock Abrasion Tool did. It is called the Sample Manipulation System, and of course your listeners cannot see this but I am showing you a photograph of this…
ROBOTS: So, it is a small [vial set space?] and then it is a circle?
What we are really studying is the habitat. Has there been or ever been signs that [Mars] was habitable for life?
Stephen Gorevan: Basically, imagine something the size of a cake – it is circular, it is a carousel. There are two rows of tubes around the periphery of the carousel, and each one of these tubes represents a sample acquisition site. What our carousel does, it is a very sophisticated machine, and what it is going to do is accept samples from… The rover has a scoop and a drill that makes cuttings; it is not like a core drill, but it makes cuttings. Or we can scoop dirt into a funnel on top of the rover. The funnel has a piezoelectric shaker; it shakes the dirt down through the funnel and into one of the 74 carousel receptacle sites, and the system tries to meter precisely about a 10th of a cubic centimeter of sample into each quartz cup. We use quartz because the cup is going to be inserted in an oven and baked to a 1000° [Celsius] and the quartz is basically neutral; it is not going to give off confusing gases for the science instrumentation. But these cups will basically take the sample that has been deposited down the funnel and move it to an oven site and then the whole carousel lifts up to press the quartz cup into the top of a chamber, and the quartz cup and the top of the chamber become an oven. But we have to have almost a hermetic seal on the top of the oven so we have to apply with a great force. And then we also come into contact with tubing when we move into this oven position, and basically we cook the sample to a 1000° [C], and gases evolve from the sample. And through an incredible series of valves and piping, that are part of a gas chromatograph mass spectrometer, an evolved gas analyser, and a tunable laser spectrometer. Each of these are instruments, and the piping network can take gases from the oven that we have brought the samples to with our device, and it could be analysed in a variety of ways using these instruments that I mentioned. This suite is called is the Sample Analysis at Mars or SAM, for short. SAM is undoubtedly the most important instrument of all the instruments on the Curiosity Rover because SAM is the instrument and we did not do this on [MERV, Mars Earth Return Vehicle?], SAM is capable of investigating organics. And we use the word organics to mean, we were asked to only say the word ‘organics’ because no one wants to take the leap to what we are really looking for here, which might be some evidence of life having been on Mars or existing. What we are really studying is the habitat. Has there been or ever been signs that [Mars] was habitable for life? And this is sort of the way this instrument suite is constructed. But going back to our role, so we are supplying this carousel which is made up of hundreds of components, it is very sophisticated, and it also has some sites that actually have cups that are covered with foil. There are six sites that are covered, the cups are covered with foil, and inside are some chemicals, and we going to direct our machine to move when the scientists determine that they want to try to look for amino acids. We are going to use this precious 6 cups, and our device is going to pierce that foil and deposit the sample inside with these chemicals and we are going to do what is called wet chemistry experiments on the surface of Mars.
And years before we were doing this project, we were in the robotics business; it is a new business; it is a new technology, relatively, in the last thirty years. To my mind it is still relatively untouched when you look at all the industry worldwide where robotics might be used and for the reasons that it might be used. One of them that interested me in the early days, years ago was the pharmaceutical industry. If one goes to a pharmaceutical company anywhere in the world and also in many universities and government laboratories there are these systems that are called Sample Preparation Systems. And basically what they are [is] what we have sent to Mars. It is an analytical laboratory system designed to do analysis on something you have made, or something you have found, results of a test, tissue, but basically these are a laboratory where you can put a specimen through a variety of precise chemical test and you can move samples down a line. We tried to automate that and we said we could go to these large – Johnson and Johnson, Pfizer, just to name a couple of companies that have sample prep systems, and probably hundreds of them worldwide – we said ‘We could make this more precise, we think, using a robot.’ And it turned out we did some; we did automate a couple of them. It turned out to be very, very complicated. But the astounding thing is that the most complete sample prep system that we ever did was not inserted in a laboratory on the Earth; it is the one we are sending to Mars.
And this is a little unusual too because usually, you want to do things on the Earth, and tried and true many times with robots, before you try it on another planet. Now we are sort skipping a step. We are sending this incredible sample preparation laboratory to Mars, and all the harsh conditions that we have, the temperature range that we have got to withstand, it is like a 100° swing, possibly, in one day. We have incredibly fine dust getting into everything, and no matter how we try to seal things this dust is going to migrate through these seals, eventually. But [what] worries the designers is about mechanisms. You are sending mechanisms to Mars. This mechanism of the SMS, three axes have movement. It is really a robotic system because it is called upon [to do] so many different tasks. It is worse in that there is going to be all that dust around, and those temperatures, not to mention rigour of the launch. The launch vibration environment is extremely violent, and so you are sending [this] complex mechanism, with hundreds of parts through that environment? No one has really done it before, not for something that operates electro-mechanically at the surface of another planet. Certainly this is brand new.
The SMS system is going to work on Mars. Probably, the plan is… (By the way I am going to personally be at the landing, which is thenight of August 5th local time in California. We will be at the Jet Propulsion Laboratory, and my job is I am part of the SAM – the name of this is instrument is SAM, as I said before – I am going to be part of the uplink team, that will send commands to SAM to do with science. The SMS is a subset of SAM. That is why I was selected to be one of the uplink people, called a PUL a Payload Uplink Lead.) …and I think that the plans are shifting as I’m speaking here. You would think that NASA has every early day planned out exactly, but they really sort of had it planned out to the first order of detail. But now we are getting closer, and we are looking at all the things we want to do in the first few sols, and we think now that in the third sol we are going to send the command to Mars to see how SAM – it is called a Live Test – and we are going to see if it survived. Of course we are all pretty confident that if the rover gets on the surface SAM will survive, but still we will be sweating bullets waiting to hear that it has.
And then, a few sols later, we will do an Electrical Baseline Test where we test to make sure all of the electrical lines on SAM that survived. And then somewhere near towards the end of August we will move the SMS, to make sure that it is functioning well, and that is going to be really a hairy time for us, because I think about all the violence of a launch, that trip through the cold of space, and the violence of the landing, and it has all these moving parts. It is just going to be very worrisome, but we will check that out near the end of the month. And then first solid sample, as it is called, the first introduction and analysis of a sample that we transport with our system, which will be a big event for the whole Curiosity payload. It is probably going to be sometime in September.
ROBOTS: Very interesting. And this is right now?
Stephen Gorevan: This is right now, and just as we did with MER [Mars Exploration Rovers?], one of the aspects that I enjoyed more than I thought [I would] was living on Mars time. It was discovered during the Pathfinder mission, which was a technology demonstrator, really, that was sent to Mars in 1996, where we saw our first American rover, Sojourner, which only went 30 yards, in travel. I have forgotten the number of how many miles we have gone and kilometres we have gone with Opportunity and Spirit, but we have come a long way and I expect…
ROBOTS: Yes, the fact that you measure it in miles rather than yards will tell us that it is there is…
So with MER it was decided we were going to live on Mars time and forget about California time, forget about it.
Stephen Gorevan: But the Pathfinder folks learned, they tried to come to work everyday living in a 9-to-5 world, but you see the rover has to also live in a 9-to-5 world, because it is a solar-powered mission, basically, or daylight mission. It gets up in the morning, does its thing, and sends its telemetry home in the afternoon. But the problem is that the Mars day and the Earth day are out of whack by 39 minutes. If you land and it is morning on the Earth and it is morning on Mars, that is fine for a few days, but eventually the Earth day is advancing with respect to Mars, and it turned out that the engineers who were working Pathfinder who tried to live on California time and tried to work within the rovers time, they got it all out of whack. Their diurnal circles got all screwed up. I mean they physically suffered more than the normal stress of the mission, and there was almost a strike after thirty days, because these guys were so affected by the disruption of their diurnal systems. So with MER it was decided we were going to live on Mars time and forget about California time, forget about it. We were given watches, some of us, that strictly work on Mars time, so in the morning you look at your Mars watch to see that your time it is on Mars and I literally forgot what time it was outside. At first they wanted to blacken our hotel windows. The windows are blackened in the control rooms. They do not want us to see the sunlight and the darkness outside they just want us to think about keeping the blinders on. But I forgot, even though I did see sunlight, I forgot local time. Although I did wear another watch in case I had to check to call my kids or something. But I didn’t prefer that watch much. So I enjoyed that, I don’t know, I just got a kick out of it, and we are going to do that again for the first 90 sols on Mars of Curiosity – it is decided that it is going to be much more advantageous – and we are going to be on-site working in Pasadena, and staying in Pasadena for the first 90 days. There will be short trips back to go home and somebody take our place for a little relief, but basically we will be there most of the time, and then after 90 sols, just as we did with MER eventually, when you learn how to control it using a sort of a bridge from Earth time to Mars time, but it is deemed that it is going to be too complicated to do that right away. There is a new payload, entirely different payload, we have got a whole set of different kinds of operations going on, and we are going to be all there doing them. But I look forward to that, in fact it is like a dream come true. I mean to be able to go to the Jet Propulsion Laboratory, and to send commands to Mars for something that you had a part in building, it is…I can die now!
ROBOTS: Oh, we hope not! Where are you going from here on?
Stephen Gorevan: Now?
ROBOTS: Well, short term I mean what is the next project, and longer term as…?
Stephen Gorevan: Well, Honeybee is maturing. We used to be overwhelmingly, after our involvement with the Space Station development in 1986, 99% of our work was for NASA, but we have become sort of approaching middle age, in terms of a company, and so we have had to branch out and not particularly put all our eggs in [the] planetary exploration basket. I think we are doing some very exciting things in other areas, and one of them was actually a fascinating story. I cannot tell you too much about it because we signed propriety agreements with the customer, but I can say we have been brought back into the private world a little bit, and one of them was in an interesting way. And that way was, despite the fact that we are a small company we have a little less than 50 people spread between three offices, we have the greatest concentration of planetary sample acquisition and drilling expertise than anywhere in the world. Of course that is an extremely narrow niche, but we are the best in the world at it. And it is fun; people here take great pride in the fact of being best at something.
People I guess read about it. I got this phone call from a woman a few years ago, and she had an Australian accent, and it turned out to be a representative of one of the great mining companies of the world. And, to make a long story short, she said, ‘Look you are involved in putting a rover in a hostile environment and taking samples, having them analyzed in situ, and radioing back results. We want to do the same thing in the outback of Australia, a hostile environment. We cannot get people to go out into these environments with the degrees that are necessary to do the analysis. We are going to, with you, build a remote vehicle to operate in the outback to drill, to take samples, and to have them analysed, and the results sent instead of to Earth, they are going to be sent back a central location in Perth, where we will have the experts who will be able to interpret the analyses that we are getting.’ So we have become sucked back into the private sector, but I am enjoying it because I never thought really that we would have this, that all the learning on Mars, with our preparation work for sampling and sample acquisition, well actually I just did not think that an Earthly application of what we were doing there would come up, but it has.
So that is one of the exciting things that we are doing in the private sector, and in the space sector we have also become involved in the Orion Project. Orion is the new deep space manned vehicle development that the United States is developing. In fact, not many people know it but we are spending two billion dollars a year to develop this spacecraft.
ROBOTS: And when you are talking deep space, is that going to the moon or Mars or further?
Stephen Gorevan: Oh, certainly just leaving Earth’s orbit.
ROBOTS: The shuttle could not do that so you need…?
Stephen Gorevan: No, the shuttle could not do that, and this would be a vehicle for exploring what we would call deep space. This vehicle with its plans for adding supplementary supply vehicles in its rear, it could go to Mars, it could go to the moon. One of the early and maybe the earliest application, the mission has not really been selected for this vehicle yet, is to go to an asteroid, and rendezvous and perhaps land astronauts on an asteroid. So it is being developed. The primary contractor is Lockheed Martin, and we are developing slip rings for transmitting power across the interface for the all-important solar arrays on this spacecraft, and we are also doing some bearings, supplying some bearing systems, that help these solar panels continuously move and face the sun at a right angle. What excites me about this is we have been around this space program for a long time. We have a great reputation, but to not get involved in a manned spacecraft where the quality control is off the charts, nothing can go wrong.
They are looking to us to join the team, because we [have] graduated, if you will, from the robotic spacecraft to the manned spacecraft, so I am very excited about participating in that project. And by the way when I say slip rings — we are robotic experts, but we are robotics hardware experts mostly, not software, although we do plenty of software here, but we have come up with ways to make robotic hardware new and innovative — slip rings allow, basically, it is a term for allowing, passing power and data across a continuously rotating interface, and the reason we became sort of experts at this is because we have been working for decades now on developing wrists of robots that can continually roll so that they might do everything from screw-driving functions, where a human has to rotate 180° at the most to operate a manual screw driver, but we get to play God in a way and develop a wrist that can continually roll. But while you still have to transfer power and data across that, if you have to continually rotating interface, you are going to have all your electrical lines wind up and be a problem. Slip rings allow you to have a continuous motion, and this is an interesting development. We have sort of been recognized for our robotics on a component level, here, and we are getting this kind of work.
Similarly we have just finished a job for the Taiwanese space agency for hinges for solar panels, and these are the kind of middle aged company space projects that we do. But some of the more advanced, more new type of jobs we are still involved in, and I think the two that come to mind are that we have invented two things for CubeSats. Now CubeSats are basically micro-spacecraft; that is becoming hot new thing now. And these are spacecraft that are literally the size of 20 cm on a side, a cube 20 cm on a side, that small. We are looking at being able to pack a lot of instrumental punch in packages that size. And the idea would be that you are sending up a larger spacecraft you can fill it with a lot of excess space – remember I said before space is more important than mass – we can put up a lot of extra spacecraft in a payload. The Defense Department has become interested in it, NASA has become interested in CubeSats, and private enterprise has become interested in cubeSats, and so we even invented and are continuing to develop our miniature, very small Control Moment Gyroscopes that are compatible with CubeSats. And also we have invented the world’s smallest solar panels for a spacecraft; they are just funny to see they are just so tiny. We have really moved down into the small sphere. So, we are very excited about becoming involved in the CubeSat wave, and perhaps the newest and the most radical thing we are working on now is what is called the Phoenix program. The Phoenix program is a brand new effort being sponsored by DARPA. DARPA is the Defence Advanced Research Project [Agency]. They are supposedly the government agency that really did give us the Internet and countless other amazing things, and their charter is basically ‘We want to develop something fast, radical and make it work.’ And I just love that, that kind of a notion. We have done some small research for DARPA over the years but this, we finally hit the big time. And what the Phoenix program is, is to finally, I mentioned it before in the 80’s, that the notion of servicing satellites, moving them to different orbits, perhaps catching and dispensing debris on orbit, it looks like this could be the really great opportunity to actually get a foothold in this area. Because my feeling is that, and I am sure the other people who are involved share it, we just need to show that we can do this. It is one of the peculiarities of the space business, to do an innovation you get to come in front of program managers and say, ‘We want to fly this.’ And then the programme managers say, ‘Where have you flown this before?’ Well, of course if you are working on something brand new like we do here the answer is we have not flown it before. So the technology that gets into space is done only very incrementally.
it was essentially not more than a whiteboard sketch when it was decided this will be on board
NASA very rarely takes a leap and flies brand new technologies. If they have, when something new appears on orbit, it usually has what is called a very high Technology Readiness Level on Earth, and that level is at minimum – and we have assigned numbers to these things – it is a minimum of 6, before NASA will even consider flying it. Interestingly the Rock Abrasion Tool is a very, very rare case of – just parenthetically here, the Rock Abrasion Tool was decided to become a payload element at a TRL level of 2, which means it was essentially not more than a whiteboard sketch when it was decided this will be on board. That is very, very, very, very rare, and it was probably one of the reasons why that not many knew about the RAT [who were not involved], because it was not in the forefront of all technologies. We were way, way, way behind, but it was decided ‘We want it’, the TRL 2 thing flown. I can’t think of another example of something at TRL 2 was able to fly. But, with DARPA behind you, DARPA is in a way fearless, they have got the kind of clout, and track record, and smart people to say, ‘We are going to get over this hump. We are going to bring technologies that do not have track records in space and we are going to get them all together and we going to fly a demonstrator, and basically a spacecraft has robotic arms on it, and it is going to go out and refuel, salvage…’ Its direct task set is not exactly set, but it is basically a satellite servicer and perhaps a debris collector. It is finally going to happen and Honeybee, again, has the honor of really developing much of the end-of-arm tooling on this device and I am very, very, very excited about that.
ROBOTS: Thank you very much for being part of the podcast.