I've mentioned the Homebrew Robotics Club here a few times. The club has an active mailing list. And when I found myself writing a lengthy post there over the weekend, I figured it might be of interest to the wider audience.
The post was in response to this inquiry from club president Wayne C. Gramlich, included here with his permission:
Can anybody point me a book that goes into design issues associated with assembling mechanisms out of bearings, axles, and gears? I'm looking for pretty basic stuff, like when to use a ball bearing, where to place bearings on a shaft, how to attach things to a shaft, etc. I am not interested in a book that tells me how to design a gear (or bearing), I just want to purchase those off the shelf.
My reply, edited a little for this post:
Wayne,
I have several different kinds of recommendations. All of these are books I own and have used.
Sometimes the best way to learn about a general topic is to see it applied to a specific purpose. So I'd like to start by pointing you to a couple of books aimed at race-car engineers and written by one of the best, Carroll Smith. These are the most pragmatic of all the books I'll be describing and contain a lot of information and advice I've just never seen anywhere else.
"Engineer to Win"
Carroll Smith
A great book. Pretty much a complete introduction to materials and structures for race car design. Most of it is highly relevant to robotics. A little dated (from 1984) but very little has changed since then that matters to us here. Even the few chapters on hydraulics, running a racing team, etc. will help put you in the right frame of mind for designing, building, and operating robots.
"Carroll Smith's Nuts, Bolts, Fasteners and Plumbing Handbook"
Carroll Smith
Another great book. Starts with the atomic structure of metals and proceeds quickly to a lot of very specific advice about selecting and using fasteners, including how to design structures that accept fasteners. Basically expands on three chapters on fasteners in "Engineer to Win," but I'm very glad I have both books.
I have a number of other books that provide overviews of mechanical design for specific applications. (Most of these are older books; I used to go to a lot of library used-book sales.) Here are a couple of representative examples, but I wouldn't recommend them unless you have an interest in these applications.
"Modern Marine Engineer's Manual"
Alan Osbourne
"Aircraft Layout and Detail Design"
Newton H. Anderson
Among the books I have that aren't aimed at specific applications, these stand out:
"Machinery's Handbook"
Erik Oberg et al.
A classic book, for good reason. Frequently updated, now on the 28th edition. I have the 22nd and 26th editions. A dense collection of data on literally thousands of topics related to machine design and manufacturing. Not one of the first books anyone should buy when getting into mechanical engineering, but certainly in the top 10. The current version is always a little pricey (currently $65.79 at Amazon) but slightly older editions are much cheaper and fine for most purposes unless you're pushing the state of the art.
There are also CD-ROM editions and a couple of companion books:
"Machinery's Handbook Guide to the Use of Tables and Formulas"
Erik Oberg et al.
Basically a book on how to use the handbook! Recommended.
Machinery's Handbook Pocket Companion
Richard Pohanish and Christoper McCauley
A shorter version of the handbook. I wouldn't bother.
"Mechanisms and Mechanical Devices Sourcebook, Fourth Edition"
Neil Sclater and Nicholas Chironis
I have the third edition of this book. It's basically just a big collection of mechanisms. Need to convert a circular motion into an elliptical motion? Select and apply an air spring? Assemble a sliding gear onto a fixed shaft? Over 2,000 diagrams are provided. The downside: the book often provides little more than a diagram and a few sentences of explanation.
"Detailed Mechanical Design: A Practical Guide"
James G. Skakoon
This book is written for mechanical engineers and does contain some formulas (which means it's somewhat pricey). But, in fact, you don't need to be a mechanical engineer or use the formulas at all because Skakoon explains what everything means in practical terms.
For example, he gives the equation for deflection of a simply supported circular plate:
| y = | -3 * q * a^4 * (5 + v) * (1 - v²) 16 * E * t³ * (1 + v) |
Complicated, sure, but Skakoon boils this formula down to the critical facts: that deflection is inversely proportional to thickness cubed and radius (or diameter) to the fourth power, and that a simply supported plate deflects four times more than a plate with fixed edges. And he reminds us that these exponential relationships mean that apparently minor manufacturing deviations can have substantial effects.
"Engineering Formulas"
Kurt Geick and Reiner Geick
This is a handy and super-dense little book. It covers all kinds of stuff, including math from arithmetic to calculus and differential equations, mechanical engineering, thermodynamics, electrical engineering, chemistry, and optics. It's very figure-heavy too, which makes the formulas easier to understand.
In terms of bang for the buck, your best bet are these free books from Stock Drive Products-Sterling Instrument (SDP-SI):
Frank Buchsbaum et al.
Available as free PDF downloads at that link, and I see someone's selling a hardcopy on Amazon, too. I have a 1983-vintage paperback edition. This book is also aimed at mechanical engineers, and also full of formulas, but it's still reasonably accessible with plenty of diagrams and explanations. In some cases the information is oriented toward the company's products, but mostly it's just good design advice. I mention the print edition because it actually contains more information than the downloadable PDFs, and the figures look a lot better than those in the PDFs. Here are some other free publications from the same company, mostly catalogs, but some include selection guides and occasional application notes that come in handy too: "Handbook Of Inch Drive Components"
"Handbook Of Metric Drive Components"
Various supplements, CDs, and catalogs
Finally, I recommend the works of Henry Petroski. He won't tell you how to design things, but he'll show you how people have designed things, which can really help you understand the design process, and especially, how it can go wrong. These are all good: "Success through Failure: The Paradox of Design"
"To Engineer Is Human: The Role of Failure in Successful Design"
"Invention by Design; How Engineers Get from Thought to Thing"
"Design Paradigms: Case Histories of Error and Judgment in Engineering"
On Thursday afternoon I was back at the Computer History Museum. The Honda Research Institute was hosting its tenth Technical Horizon Symposium and announcing this year's Honda Initiation Grant awards.
Honda's Asimo robot
(Credit: Peter Glaskowsky)The grants are part of the Institute's efforts to stimulate collaborate research between Honda and the academic community. Since 1997, Honda says it has awarded 75 grants totalling "several million dollars" to universities in the US. This year, Honda received 300 proposals; it chose seven. This year's awards (listed here along with those of past years) cover research in safety, efficiency, emissions control, and user interfaces.
Also on hand for the event, which attracted an audience of some 300 people, was Asimo, Honda's famous robot. This Asimo is actually the second-generation model, and there were also three generations of prototypes. Over the years, Honda's been able to reduce the size of the necessary motors, power supplies, and control systems; the current Asimo is a cute little thing, just 4'3" (130cm) tall. Although it conveys the impression of solidity and weight, it's actually just 119 pounds (54kg). At this size, Asimo is big enough to interact with humans without posing much of a threat in case it bumps into someone or--as it can do if the power fails suddenly--falls down.
Honda put Asimo through its paces for us-- walking around the stage, balancing on one leg, kicking a soccer ball, traversing a set of stairs, and even running. The latter skill involves a peculiar loping gait; it's almost impossible to tell that Asimo is actually running, but Honda assures us that both feet leave the floor for about 80 milliseconds, during which time the robot moves about 2 inches forward. I was surprised to learn that Asimo is controlled by just four microprocessors, only two of which manage balance and locomotion.
In the audience were several members of the local Homebrew Robotics Club and the founders of Anybots, which I wrote about back in September. Asimo is far beyond the accomplishments of Anybots and other developers of autonomous robots... but then, it should be; Honda has poured untold amounts into its development. I'd guess the total amount must be in excess of $40 million, but Honda isn't saying.
Anyway, it was interesting to get a close look at Asimo. It's an impressive accomplishment, but it has a long way to go before it's ready for commercial sale. I suspect Honda's investment to date is just a drop in the bucket compared to the work that still remains. I can't begin to guess whether Honda will ever recoup its investments, but I'm glad it's doing the work.
You've probably seen or heard of the industrial robots that build cars, and the various humanoid robots like Honda's Asimo. Most of these are made in Japan. But let's face it, there's only so much these can do. An industrial robot is bolted down, and only knows one or two simple tasks. Asimo is small and weak, and famously collapsed once while trying to climb stairs.
Monty, a wheeled robot from Anybots.
(Credit: Anybots, Inc.)As we know from sci-fi movies, real robots are the size of a man and can do things--dangerous things. Real robots are suitable for building robot armies. For that, we have to look to America. Companies in the United States have given us vacuum-cleaner robots, bomb-disposal robots and even robot military aircraft. Where better to look for humanoid killer robots?
And I found some on Wednesday night at the headquarters of Anybots, a six-year-old angel-funded Silicon Valley start-up company of just four employees.
Of course, Anybots doesn't talk about robot armies. Its robots can't kill anyone, not even by falling on them--they're not heavy enough for that. In fact, the company's favorite publicity photo shows one of its robots loading a dishwasher. But I can tell they're really thinking about robot armies. I've seen this before...in bad sci-fi movies, anyway, but now I think someone's trying it in real life.
The mad scientists of Anybots, I believe, are just lulling us into a false sense of complacency with these pleasant domestic demonstrations.
We learned on Wednesday that Anybots is planning to make its fortune from these peaceful applications. In fact, the Anybots employees we met seemed to be good people, sincerely devoted to this idea that robots can free mankind of dishwasher-loading and similar drudgery. If any of them was an evil genius, he (or she) escaped my detection. Well, maybe I'm just paranoid. As you may have heard, that's a useful survival trait here in Silicon Valley...
Anybots was hosting a special monthly meeting of the Homebrew Robotics Club, which I've mentioned here previously. We got a presentation by Anybots founder and CEO Trevor Blackwell, who summarized the company's vision quite efficiently: "Make robots that do what people can do."
Initially, Anybots is developing teleoperated robots--that is, robots operated by a person some distance away. Autonomous robots, which operate independently, are further in the future. The major challenges of humanoid robot design are the same for both cases: balancing, moving, recognizing objects, aiming rifles and conquering the world. (That's just me talking; Blackwell said nothing about military combat. Clever man.)
We also got to see a couple of Anybots' robots. Monty is the one pictured here and on the main page of the Anybots Web site. It has a two-wheeled balancing base; the balancing function is autonomous, so the operator doesn't have to worry about it. Unlike Honda's squat Asimo, Monty is 5'7" tall, so it has the height and reach of a (small) man. On the other hand, Asimo is self-powered, whereas Monty must be connected to an external source of compressed air and electrical power.
According to Blackwell, Monty can basically go anywhere a wheelchair can go. It's powered by a combination of electric and pneumatic actuators, which has proven to be a challenging design decision. As Blackwell said, few other companies use pneumatics for robots, and Anybots is learning why...
The other robot is Dexter, which is a few inches taller and walks on its own two legs; it's also permanently tethered to its air and power supplies. It even wears human shoes, and walks using an algorithm that models the way humans walk. (However, the poor thing has no arms, and when we saw him, he didn't even have his head on!) Dexter is still learning to walk, but Anybots expects it will eventually be able to run, cross rough terrain, and generally transport itself anywhere a human could go.
Anybots is a sponsor of the RoboDevelopment Conference and Exposition, coming to the San Jose McEnery Convention Center October 25-26, so you can go there to learn more about Anybots' work and other robot-related technology.
It might be a good idea to go check out the show just so you can figure out which side you want to be on in the inevitable war between man and robot...
Have you ever heard of the Homebrew Computer Club? I'm sure you've heard of the products designed by its members: the Apple I and Apple II, the Osborne I, maybe even the earlier Sol-20 (one of the prettiest little personal computers ever; I have a beautiful example myself).
Wikipedia reports that the Homebrew Computer Club stopped meeting in "roughly 1977"-- about 30 years ago. But a small part of it survives. Some of the people in the Homebrew Computer Club spun off the Homebrew Robotics Club, and that club still meets regularly.
I try to attend meetings when I can, but I've been missing a lot of meetings since I took this job at Montalvo Systems. I missed the meeting this month; blogging about it sorta helps make up for that.
HBRC members are still mostly engineers and programmers. Some are parents and kids, which bodes well for the long-term health of the hobby as well as the long-term supply of workers for high-tech industries. Club projects involve everything from simple wheeled robots that can drive around on a tabletop without falling off the edge to GPS-equipped machines like small versions of the DARPA Grand Challenge vehicles. Pretty much every robot has a microprocessor brain; some have many microprocessors, each in charge of some subsystem.
The connection between microprocessors and robotics is not mere coincidence. Microprocessors revolutionized robot design. It was certainly a delayed effect, since microprocessors had to grow up for a while before they surpassed the previous methods for robot control. But when this happened, more or less in the 1990s, microprocessors put robots on the Moore's Law path.
No longer must robots remain tethered to industrial minicomputers, slaving away day after day to build Honda automobiles. In fact, Honda itself is making robots that can walk off the job if they want to.
Of course, they're not smart enough to want to, which brings me to the final ingredient required for intelligent, independent robots: synthetic brains. PC Magazine and CNET covered the recent Cognitive Computing conference, where researchers described projects to reimplement the brain in silicon. Not merely duplicate the results of thought, but the process itself.
I attended a presentation on this subject by Jim Burr at the 1993 World Science Fiction Convention in San Francisco. If I recall correctly, Burr projected that it should become possible to implement an electronic copy of the human brain, in a size that would fit inside a human head, consuming a similar amount of power, by 2050... and built in silicon, which switches far faster than biological neurons, the electronic version will be a thousand times smarter.
Moore's Law has probably slowed down a little since then, and it might slow down more, but it's still possible some of us will live to see this achievement. What happens then, I can't say, but maybe robots with silicon brains will start an Organic Humans Club...
- prev
- 1
- next
