Date: Tue, 30 Apr 1991 19:51-EDT From: space-tech-request@cs.cmu.edu To: ~/st/lists/stdigest Subject: Space-tech Digest #79 Sender: mnr@DAISY.LEARNING.CS.CMU.EDU Contents: Marc Ringuette Location of space-tech archives Marc Ringuette Satellite repair robots [ from space-investors ] ------------------------------------------------------------ Date: Fri, 26 Apr 1991 18:02-EDT From: Marc.Ringuette@DAISY.LEARNING.CS.CMU.EDU To: space-tech@cs.cmu.edu Subject: space-tech intro I archive the space-tech digests & excerpts on daisy.learning.cs.cmu.edu (128.2.218.26) in directory /usr/anon/public/space-tech. You can use anonymous ftp, or you can send me mail at space-tech-request@cs.cmu.edu and I'll send you what you need. The excerpts are 150k of selected stuff: EM Launchers Tether Propulsion CMU Mars Rover High Velocity Guns Mars Mission Ship Design Orbital Debris Launch Loops Robots on the moon The digests contain the full 1.5 meg of space-tech postings. Feel free to start something up, by the way! ----------------- -------------------------- --------------------------- | Marc Ringuette | Cucumber Science Dept. | Never lick a gift horse | | mnr@cs.cmu.edu | Cranberry Melon Univ. | in the mouth. | | 412-268-3728 | Pittsburgh, PA 15213 | | ----------------- -------------------------- --------------------------- ------------------------------ Date: Fri, 26 Apr 1991 18:30-EDT From: Marc.Ringuette@DAISY.LEARNING.CS.CMU.EDU To: space-tech@cs.cmu.edu Subject: Satellite repair robots Here's some discussion from Vince Cate's new space-investors list (space-investors-request@cs.cmu.edu) about satellite repair robots. Feel free to continue it here or join that list. ------------------------------ Date: 25 Apr 91 09:36:03 From: Michael Wallis Subject: RE>Re- RE>Et Tu Space News? To: space-investors@cs.cmu.edu As the ANIK E-2 is showing, yes, a stuck antenna CAN make a perfectly good launch a complete write-off. ... A perfectly good satellite is going to sit out at GEO as a write-off becuase of a stuck antenna when 20 minutes with a wrench would free it. A manned flight out, a quick fix, and then home to file the salvage claim. The whole thing would cost under $50 million for a $100+ million satellite. Not a bad deal. Michael ------------------------------ Date: Thu, 25 Apr 1991 14:49-EDT From: Marc.Ringuette@DAISY.LEARNING.CS.CMU.EDU To: space-investors@cs.cmu.edu Subject: Salvaging a satellite > A manned flight out, a quick fix, and then > home to file the salvage claim. The whole thing would cost under $50 million > for a $100+ million satellite. Not a bad deal. Not a terrible deal, but wouldn't it be better to send a little teleoperated robot? I daydreamed last year about building such a little robot (maybe 50-100lbs) with a camera and a couple of arms, electric propulsion, and maybe a drill and some krazy glue. You'd put it up in GEO and wait for the next satellite to get stuck, then give them a call. Potential thousand times return on investment. ----------------- -------------------------- --------------------------- | Marc Ringuette | Cucumber Science Dept. | Never lick a gift horse | | mnr@cs.cmu.edu | Cranberry Melon Univ. | in the mouth. | | 412-268-3728 | Pittsburgh, PA 15213 | | ----------------- -------------------------- --------------------------- ------------------------------ To: space-investors@cs.cmu.edu Cc: yamauchi@cs.rochester.edu Subject: Repair Bots Date: Thu, 25 Apr 91 15:38:11 -0400 From: yamauchi@cs.rochester.edu Michael writes: >I sure wish I had the SSTO >vehicles I've bee trying to get funding for. A prefectly good satellite is >going to sit out at GEO as a write-off becuase of a stuck antenna when 20 >minutes with a wrench would free it. A manned flight out, a quick fix, and then >home to file the salvage claim. The whole thing would cost under $50 million >for a $100+ million satellite. Not a bad deal. Here's an idea I've been thinking about: It seems that many of the satellite problems are caused by relatively easy-to-fix failures (antenna/solar panel deployment failure, swapable module failure, etc.). Perhaps these could be fixed by a small (disposable) repair robot that could be launched on a booster like Pegasus. In order to make this economically feasible, the core design would have to be standardized, but it would also need the ability to have small custom modifications made for each task. For example: a robot with two six degree-of-freedom arms, one of with a dextrous multifingered hand, the other capable of mounting interchangeable specialized end effectors designed for the specific application, stereo cameras and/or laser rangefinder, force/torque sensors, and thrusters for maneuvering. Basically, you could look at this as either a small, disposable EVA Retriever or a brilliant pebble with arms. This could either be teleoperated or autonomous. The best bet might be some sort of hybrid. Autonomous behaviors could be used for actions which are simple and require high-precision or low-latency. For example: protective reflexes (collision avoidance, force/contact sensing), simple high precision manipulation (unscrewing a fastener after the driver has been placed next to the screw), and maneuvering to rendezvous. Human teleoperation could be used for tasks requiring complex sensory processing (what is blocking the antenna from deploying?), complex manipulation, or reaction to unforeseen circumstances. As to how much this would cost: a large Puma arm costs under $100,000, a Utah/MIT Dextrous Hand plus control computers costs around $250,000, a two-degree of freedom stereo camera platform costs in the low tens of thousands, I seem to remember hearing that the price for a high-quality laser rangefinders was around $100,000, and you can buy a high-power workstation for a price in the low tens of thousands. So this would come in at under $1 million for an earth-based system. Multiply this by a factor of 10 for integration, space-qualification, and limited production runs, and you have a cost of around $10 million. If this is going to be teleoperated or hybrid add, say, $3 million for a communication system (does this sound reasonable?). Add another $1 million for whatever simple custom tools (wrenches, gripper attachments, etc.) need to be machined for each mission. I have no idea how much a thruster system would cost, but it's important to remember that this is a disposable system, so it does not need the reusability of an MMU or the EVA Retriever's maneuvering system. I'll throw in a wild guess of $6 million -- if anyone has a better guess, I'd be interested to hear it. Add in another $5 million for a Pegasus launch, and the total cost comes to $25 million. Sounds like a pretty good investment to save a $100 million satellite. One point to remember on space qualification -- this doesn't mean mil-spec or NASA-spec. It just needs to have a fairly high probability, say 90%, of working long enough to complete it's mission. Even if it takes two launches (with a 99% chance of getting the job done), you'll have saved $50 million. Any comments, either technical or financial? I'm more confident about the robotics aspects (since that's my field) than the prices (which are just semi-educated guesses). Brian ------------------------------ Date: 25 Apr 91 11:52:32 From: Michael Wallis Subject: RE>Salvaging a satellite To: space-investors@cs.cmu.edu Reply to: RE>Salvaging a satellite > Not a terrible deal, but wouldn't it be better to send a little > teleoperated robot? Possibly, but Skylab and Shuttle have shown the advantage of having people there for repairs. Besides, I'm a hands on kind of guy. It'd rather be there to decide what needs to be done than hope I sent up the right equipment and enough fuel to apply the force needed to free whatever is stuck (assuming what's stuck is where a robot can get to it without damaging or contaminating the prize). Further reflection at lunch has lead me to believe it could be done for considerably less than the $50 million quoted earlier. Using the Alpha model outfitted with a lifesupport module in the payload bay and carrying provisions for a 72 hour mission, refueling on orbit, the whole thing could be done for about $5 million; a MUCH better cost on a $100+ million satellite. What's interesting also is that I've been reading some of the space/commercial thread on BIX (Byte Information Exchange) from last October (when I was last on regulary) and they were discussing whether or not the market was elastic with respect to price (ie if the price to orbit dropped by an order of magnitude would there be enough buyers to keep a launch company going). I think this type of service (on orbit repair or salvage) could be a real profit maker. There's also retrieval of old satellites now on orbit that are either refurbishable and can be resold, or not salvageable and are a hazard that needs to be removed. Lots of opportunity to make a buck. ------------------------------ Date: 25 Apr 91 12:35:10 From: Michael Wallis Subject: RE>Repair Bots To: space-investors@cs.cmu.edu Several comments. First, $25 million (our figure) is a lot to 'throw away' each time. Admittedly, it's less than the insured value by an order of magnitude but it's still a oneshot. Second ... I think you're underestimating the complexity of space docking. The Soviets have had a LOT of practice at this and they still do a significant number of their dockings by hand. Their manned flights are almost always by hand and their Progress runs have a bad habit of taking several passes before they make it in safely. They missed pranging MIR with the Kvant-2 module by only 10 meters! You wouldn't want to hit the satellite you were trying to 'rescue' . Thirdly, operating repairs in a micro-gravity environment is a tricky thing at best. I'm not sure your repair-bot would be able to provide the torque needed without either damaging or contaminating the satellite you're trying to save. And yes, you WILL need rad-hardened electronic componants. GEO is well outside the Van Allen belts and you're going to have to be thermal and electromagnetically insulated. Michael ------------------------------ From: Nick Szabo Message-Id: <9104260113.AA14865@crg5.sequent.com> Subject: Re: Repair Bots To: space-investors@cs.cmu.edu Date: Thu, 25 Apr 91 18:13:03 PDT The satellite repair robot design sounds quite interesting. The most important missing link is the satellite design. Contemporary satellites are not designed to be repaired, by either human or robot. We need to research what % of failures are external or otherwise easily accessible, like antennas being stuck. As a WAG, perhaps the market for this is between $20 million/year (one fix every five years) and $100 million/year (one fix/year). If we can keep R&D costs low like Pegasus, the robot discussed pays off at the high end of that range. If we can convince new satellite builders to redesign future spacecraft for service by our robots, the market greatly increases -- perhaps to between $100 million and $300 million per year. If so, we have the NPV of a cash flow between ($80 and $200 million - expenses) per year -- say $40 million/year => $300 million -- to spend on satellite redesign. The business problem is forming a joint project with satellite manfucturers such as Hughes and General Electric to finance the redesigns, with revenue for the robot repairs being shared between the repair company and the satellite builders. As suggested on sci.space, insurance companies can play a key role in bringing the repair company and satellite makers together and financing the project, since it is the insurance companies that pay the c. $100 million and end up owning a fleet of useless satellites. As a side note, this is similar to the business problem of convincing upper stage and satellite makers to design for refueling, given a set of extraterrestrial resources processed into fuels in low earth and/or geosynchronous orbits. A further market to study is the possibility of reviving satellites that have failed in previous years. Depending on the rate of fuel leakage, radiation damage, and other factors, there may be a supply of several hundred million dollars' worth of dead circuits sitting in GEO. If so, this alone could finance the development of a robot sophisticated enough to help revive these satellites. The last comment is a repeat -- for Pegasus to launch a robot to GEO, it needs an upper stage. That in itself is an excellent business opportunity and important for further growth in the entry-level commercial market. Nick Szabo szabo@sequent.com "Living below your means allows you to live better than living above your means." -- Dave Boyd The above opinions are my own and not related to those of any organization I may be affiliated with. ------------------------------ From: Nick Szabo Message-Id: <9104260243.AA19413@crg5.sequent.com> Subject: satellite repair vehicle To: space-investors@cs.cmu.edu Date: Thu, 25 Apr 91 19:43:03 PDT > First, $25 million (our figure) is a lot to 'throw away' each time. How much is a "lot"? I would say that $2,000 million+ for development of a manned upper stage, which would throw away far more mass and engine work than we put into our robot, is a lot for a risky market between $20-$500 million/year. We are talking an NPV in the range of $100 million (for the repair company side) here. $2,000 million is a factor of 20 too high. Now is $25 million a lot compared with $100 million in revenue? That's a 400% gross margin -- and venture capitalists start drooling at 50%. As long as the initial R&D costs can be kept in the $25-$100 million range, and the market pans out -- we need further research on both of these assumptions -- we have one heck of a profitable business here. > Second...I think you're underestimating the complexity of space docking. The > Soviets have had a LOT of practice at this and they still do a significant > number of their dockings by hand. The Soviets also stand in long lines for potatoes and foreign-made 8086's. They are third-world where automation is concerned. Space docking is little more than normal spacecraft control, knowing the geometry and location of the target, and computation. The U.S. has had all these technologies, and their combination in automated docking, well under control since the mid-60's. As a side note, the cosmonauts nearly got themselves killed recently when they unknowingly bumped into the Progress docking antenna during an EVA. Having people on-site has its problems as well as its benefits. > Thirdly, operating repairs in a micro-gravity environment is a tricky thing at > best. I'm not sure your repair-bot would be able to provide the torque needed > without either damaging or contaminating the satellite you're trying to save. Good point. A critical part of each mission is thoroughly mapping the satellite blueprints, designing the torque-holds accordingly, and testing out the holds on earth-based models of the satellite and robot, perhaps in a swimming pool for boyancy micrograv simulation. We need a garage version of the SOP for Shuttle-based repairs. > And yes, you WILL need rad-hardened electronic componants. GEO is well >outside the Van Allen belts and you're going to have to be thermal and > electromagnetically insulated. Another good point. The robot will have to be designed with electronics standard for GEO satellites, not earth-based robots. The robot arms and other vacuum-exposed moving parts must be designed to avoid vacuum-welding -- fairly tricky and non-standard stuff we can borrow from the Viking lander arm, the lunar rover and/or the Shuttle arm. The software will have to be squeezed into what we consider fairly primitive electronics -- a customizable silicon compiler will be useful here. All the technology pieces are being done now in space and on earth, and it will take a fair amount of R&D -- software development and testing, hardware design and prototyping, vaccuum and radiation chamber tests, test flights, etc. -- to bring it to life in combination. If we follow the get-hands-dirty techniques used by the Pegasus team, I suspect that we can keep those costs in the $25-$100 million range. But your point is well taken -- an important part of the design is the painful reduction step of substituting space-proven components for earth-proven components, and designing new hardened components where needed. As a side, note, from knowing the details of particular missions we can consider designs that go beyond anthropomorphic concepts of a robot. For example, a simple lasso might be sufficient for freeing a stuck antenna. A significant number of satellite repairs may be possible with surprisingly simple designs, even designs customized for single $100 million missions. With modern tools such as CAD, stereo lithography, and computer-controlled machine tools, several alternatives can be tested in the swimming pool and the best used for each mission. A detailed look at the designs of the dead satellites up there right now, and what fix(es) are needed to revive them, is very beneficial both for market research and the actual design of the repair vehicle. Nick Szabo szabo@sequent.com ------------------------------ From: Nick Szabo Message-Id: <9104260327.AA20877@crg5.sequent.com> Subject: The cost of GEO astronaut capability To: space-investors@cs.cmu.edu Date: Thu, 25 Apr 91 20:27:46 PDT > Using the Alpha model > outfitted with a lifesupport module in the payload bay and carrying provisions > for a 72 hour mission, refueling on orbit, the whole thing could be done for > about $5 million; a MUCH better cost on a $100+ million satellite. A typical Shuttle repair costs $50 million. An aerospace rule of thumb is that things in GEO cost 4 times what they do in LEO. We have no astronaut capability in GEO. How do you propose to achieve such a capability and this incredible factor of 40 savings on an R&D budget of less than $100 million? In comparison the Shuttle, which merely goes to LEO, cost over $20,000 million in today's dollars to develop. The only U.S. system with astronaut GEO capability, Saturn/Apollo, is now unavailable but also cost over $20,000 million to develop. An even more incredible savings of a factor of 200 is needed for the R&D. This is not counting the repair equipment itself, which requires most of the same R&D (hardening, vacuum-prooofing of moving parts, boyancy micrograv simulation, etc.) we need for the the proposed robotic version. Nick Szabo szabo@sequent.com ------------------------------ From: Randell Jesup Message-Id: <9104260548.AA25450@cbmvax.cbm.commodore.com> To: space-investors@cs.cmu.edu Subject: Re: Repair Bots Another potential use of repair bots (actually, you'd want a different standard unit just for this): boosting satellites that didn't have their upper stages ignite, or ended up in the wrong orbit. even if it takes a couple of these to get it to the right orbit, that's still a big savings over a lost satellite (since these will be even cheaper than the maintenance bot's - just a booster with some clamps to grab on and guidance). Randell Jesup, Keeper of AmigaDos, Commodore Engineering. {uunet|rutgers}!cbmvax!jesup, jesup@cbmvax.commodore.com BIX: rjesup Disclaimer: Nothing I say is anything other than my personal opinion. Thus spake the Master Ninjei: "To program a million-line operating system is easy, to change a man's temperament is more difficult." (From "The Zen of Programming") ;-) ------------------------------ Date: Fri, 26 Apr 91 07:57:37 -0500 From: David Cornutt To: space-investors@cs.cmu.edu Subject: Re: repair bots Hmmm... interesting. Let me modify the proposal a bit. I have a natural aversion to throwaway space hardware, so let's make the thing reusable. (I think the costs of space-hardening are a bit underestimated anyway. As several have already pointed out, without significant rad hardening, the chances of the bot surviving launch and the trip through the Van Allen belts are significantly lower than 90%.) First, launch it into an orbit slightly above geosync, using an appropriate vehicle. (I don't know that Pegasus is up to it; it seems to me to be more of a Delta-class proposition.) Equip it with a cold-nitrogen system for maneuvering; it's much simpler than hydrazine combinations, and less likely to contaminate the target. When a repair needs to be done, wait until the target comes around under the bot's orbit, then have the bot retrofire to match orbits and rendezvous. The bot, under ground control, then looks over the situation. If parts need to be sent up, use a Pegasus to do that. The bot is then docked by remote telemetry from the ground, and the repairs effected. When this is complete, the bot fires itself back into its above-geosync orbit. Periodically, a Pegasus is sent up with more nitrogen. About the docking: it is true that the Soviets have done more of it than we have, and it is also true that they are way behind on the necessary technology. I knew some folks who were working on OMV before it was cancelled, and they had a telecontrol docking simulator that simulated the space-to-ground delays and all of the orbital mechanics. I flew it some, and was able to dock with spinning targets after a few tries. (I never succeeded with a tumbling target, but they told me that they had people who could do it consistently.) Dave David Cornutt, New Technology Inc., Huntsville, AL (205) 461-6457 (cornutt@freedom.msfc.nasa.gov; some insane route applies) "The opinions expressed herein are not necessarily those of my employer, not necessarily mine, and probably not necessary." ------------------------------ From: Nick Szabo Message-Id: <9104261640.AA05445@crg5.sequent.com> Subject: Re: Repair Bots To: space-investors@cs.cmu.edu Date: Fri, 26 Apr 91 9:40:12 PDT > Another potential use of repair bots (actually, you'd want a different > standard unit just for this): boosting satellites that didn't have their upper > stages ignite, or ended up in the wrong orbit Another good idea! This is a great forum. This idea might be implemented as a combination of 1,000 kg of extra fuel, 100 kg of extra tank, and 100 kg of mating adaptor and rendesvous electronics launch on a Delta to GTO or other near-GEO orbit where the satellite is stranded. We will need to work out the exact amount of extra fuel needed for specific missions, but our mass budget will probably be more than the c. 150 kg available from a Pegasus with upper stage to GTO. As usual, market research will be needed to see how many otherwise usable satellites are in close but useless orbits. My guess is there is a market for 4 satellites of $100 million value up front, plus 2 every 5 years, for a total of $400 million the first year plus $40 million per year thereafter. With some further business wheelings & dealings, the rendesvous technology might be shared with the previously proposed repair business, reducing the cost of R&D for that component. $40 million launch + $10 million upper stage + $10 million special stage and adaptor + $4 million insurance + $4 million operations = $68 million, for recovery of circuit value that should average $100 million. This business is more tied to launch costs limits, so the gross margin is only 47% -- on the low end of VC standards, but perhaps still reasonable. This gives a cash flow of $128 million plus $12.8 million per year, for an NPV (net present value) of $212 million. By rule of thumb this gives an R&D budget for the special stage, adaptor, and rendesvous equipment of $105 million. ------------------------------ Date: Fri, 26 Apr 1991 17:26-EDT From: Vincent.Cate@FURMINT.NECTAR.CS.CMU.EDU To: space-investors@cs.cmu.edu Subject: Re: Repair Bots Another interesting possibility is refueling satellites that are out of fuel or attaching a new positioning rocket to them. Satellites run out of fuel now and become useless after X years where X is something like 6 to 12. It is not clear to me how hard it would be to add fuel to your average satellite but my guess is that since it has to be done before they are sent up it is possible to add fuel. Anybody know? Note that, if you are not in a hurry, you can use solar cells and a high ISP drive to boost something into a high orbit. So with the existing and future satellites in need of repair (esp simple stuff like yanking on an antenna), or boosting, or a fill up, I'll bet there is enough of a market to justify this. We should try to generate a list of satellites that either have stuck antennas, did not make it to GEO, or would be useful if they had more fuel. Then maybe work out a business plan would be fun. If the startup group had an agreement from 10 places for a payment of $50 mil for each fixed satellite and a plan that needed $50 mil to get started and then $7 mil per sat after that (new fuel) I think it would be possible to raise venture capital. Some fun... -- Vince ------------------------------ From: Nick Szabo Subject: Re: Repair Bots To: space-investors@cs.cmu.edu Date: Fri, 26 Apr 91 13:51:10 PDT > Note that if you are not in a hurry you can use solar cells and a high > ISP drive to boost something into a high orbit. Hmmm, well... remember those Van Allen belts. Very nasty for solar cells to stay in them for the several weeks to months it would take to spiral through with the thrust on contemporary electric engines. The OSC/RRC Prometheus proposes a hybrid chemical/electric engine to avoid this problem. > We should try to generate a list of satellites that either have stuck > antennas, did not make it to GEO, or would work if they had more fuel. Besides the tech issues, this is the crux data needed at this point. References greatly appreciated. > If we had an agreement from 10 places for a payment of $50 mil... This is another good business strategy: get orders ahead of time. Makes financing much easier (=> more $$ for R&D). There will probably be a "credibility feedback" between the ability to get orders and get financing. Nick Szabo szabo@sequent.com ------------------------------ From: Vincent.Cate@FURMINT.NECTAR.CS.CMU.EDU To: space-investors@CS.CMU.EDU Subject: Re: Repair Bots >The OSC/RRC Prometheus proposes a hybrid chemical/electric engine >to avoid this problem. Prometheus could be a great starting point for a "Repair Bot". This could cut development costs down alot, easily to under $30 mil to get started. So you would not need to line up too many customers at $50+ mil a pop to get venture capitalists interested. I don't know how big it is but I expect it would fit into a Pegasus, at least without the fuel. Sending fuel up separate would be reasonable enough since that is what you will do in the long run anyway. So, does anybody know how big Prometheus is? Isn't NASA working on something like this? Or were they just talking about it? -- Vince ------------------------------ End of Space-tech Digest #79 *******************