Slashdot Mirror
Preparing for a Career in Robotics?
seanfast asks: "I just graduated from college with a B.S. in CompSci and a specialization in Artificial Intelligence. I am currently working full time, but I want to go back to school part time for my M.S. and specialize in AI or Robotics. Unfortunately, with my time schedule, and the extreme scarcity of a degree with either of those specializations being in my vicinity, I will most likely have to settle simply for a M.S. in CompSci with no specialization. If I want to work in the field of robotics and AI later on in life, what do I need to do in my current situation to prepare myself? Some have told me I need a strong mechanical engineering background, some have said I need a stronger software background, and some say I need to just tinker with stuff in my free time and not even worry about what they can teach you in school. Any advice, Slashdot?"
Post jokes about welcoming overlords here. Thank you.
I can't give you advice on how to get into robotics because I never successfully did that. I worked with pioneer robots mounted with laptops and had the whole Aria package figured out. I studied all the white papers and took all the courses. I'm even getting my masters with a specialization in AI. What was my problem? I'm not sure, it was probably the fact that my grades were ~3.5 GPA out of 4.0 & I've never been published.
If you really love this topic and will settle for nothing less, then you have to be prepared to devout a lot of time to reading about everything out there and, yes as you mentioned, tinkering with things like JStick and real time microboards all the time. You need to be a master of forward & inverse kinematics and also have all the algorithms down pat.
I say this because people are not ready to hand over responsibilities to robots. You might cite NASA but their rigorous protocol of checking and double checking every tiny movement of their robots anything but artificial intelligence. Reason? High failure rate otherwise.
Today's robots leave a lot to desire. That might have changed since I last looked in the field but I can tell you that less than 5% of all computer scientists are lucky enough to work with robots (or unlucky enough) and I think an even smaller percentage get to develop for them. Maintenance is just as needed there as it is in any other software.
I'm not trying to discourage you, I'm trying to be realistic. I read I, Robot in fifth grade and it changed my life. Unfortunately, it only gave me the desire, not the rigorous technical background needed to put me in the few high percentage points of students.
You mention mechanical engineering but that implies robotics from scratch. If you're a computer science student, I advise you to treat the hardware as a blackbox and use the APIs to program for them. There is some cross over you will need to learn to program for arm or walking robotics but this is more theory of how your code should look to work the controllers. I guess if you want to design from scratch and make genuinely new physical robots, then you need not only a mechanical engineering background but also one in electrical engineering. I also foresee a lot of the signals moving from hardwired to wireless for simplicity so that would mean Fourier transforms, wavelets, & the like.
My suggestion is to hit Citeseer for the free papers. Hit your college's IT site and try to get into the IEEE Computing document repository. They also have a special robotics division that you might find useful for creating contacts though I'm a member of it and that's never happened (you have to attend a lot of meetings). Look everywhere for material on the topic and see what other people did right and what other people did wrong. Have you ever heard of Robocup? Definitely read all the papers released about that and look into becoming active in your university's robotics lab.
Most importantly, keep yourself knowledgeable/marketable for conventional jobs in computer science because you really never know what's going to happen. Robotic development has a very limited market. The factory line robots are getting more and more reliable and it seems any biomimicked robotics are for purely entertainment value. I'm not intending to be mean when I say it, but there probably is no "career" solely in robotics. You've got to bus tables in the computer science world before you can prove yourself to the big dogs.
I now write web services and web applications. You have a romantic goal, I wish you the best of luck in a more exciting future.
My work here is dung.
Most of what I see out there isn't robotics like sci-fi androids, where AI is needed. Its manufacturing work, and automation of small tasks (such as robotic vaccuum cleaners and dishwashers). Such things don't really need a lot of smarts, they just need precise movement and improved ease of programming/lower cost. If you really want to program them, study up on firmware and embedded programming, and machine vision (the only real AI most robots need). A basic understanding of ME is nice, but a degree isn't necessary.
I still have more fans than freaks. WTF is wrong with you people?
You already have specialization in A.I. so I would suggest getting your foot in the door at a company that works in A.I. if you aren't already. Let them send you to school for your M.S.
Depeding where this position is located will solve the problem of just getting an M.S. in CompSci or a more specialized degree.
"A government is a body of people, usually notably ungoverned." - Shepard Book Quoting Malcolm Reynolds
Then think of one that has offices in almost every major city so this guy's got a chance applying there. I'm waiting.
If I want to work in the field of robotics and AI later on in life, what do I need to do in my current situation to prepare myself?
If that is really the only thing you want to do (i.e., life-long passion and all that), then go to CMU. AFAIK, there is no better robotics program than the one at Carnegie-Mellon.
...in fact, I did. Got accepted too.
Then a bit of further research revealed there were literally fuck all jobs in the field, and the univeristy s star Robotics student had only managed to get a job a Stationery factory. The same company that provided all of the University's stationery, I might add.
I'm now about to embark on a Degree in Conservation Biology and Ecology. There may be just as many jobs (read: zero), but at least I'll be doing something worthwhile.
Unless you are deadly serious and know you can find a niche in a sector where there are no niches, I wouldn't bother mate.
cmd-q.co.uk - some sort of stupid fucking internet bullshit
I ran one of the DARPA Grand Challenge teams, Team Overbot. I'm not sure there really are "careers in robotics". Of our best young people, one is now running a hedge fund, and the other is working for a financial derivatives firm in New York. Neither of them could find anything in robotics with a big payoff.
With 12 million illegal immigrants, the US doesn't need robots. Japan, though...
The only advice I have to offer (and I have yet to take it, so ymmv) is: Obtain and play with Lego Mindstorms. I'm sure other people will offer better advice.
If you evar decide on a plan of action, then please share.
[o]_O
"Unless you are deadly serious and know you can find a niche in a sector where there are no niches, I wouldn't bother mate."
Go work for Honda.
pretty much regardless of what you want to do in robotics (other than at the technician level), you're going to need a master's degree and quite possibly a ph.d, depending on where you end up working. part of the process of getting the advanced degree means accepting that you are probably not one of the very few people in the world who are able to master robotics from the high level control / deliberative planning, low level reactive planning, perception, mechanical and electrical control, communications & processing architecture, all the way down to the low level circuits and mechanical systems. so pick something you want to do well in (in my case it's the machine perception and sensing side) and write a master's thesis (or do a master's project) on something relating to that speciality.
:-)
regardless of your chosen speciality, you're going to want to get to be really good at math. in particular, linear algebra, calculus (multivariate), trig, analytic geometry, and stats/probabilities (in particular, bayesian thinking). also make sure you are really solid on forward and inverse kinematics. finally, an understanding of communication systems will be incredibly useful.
concurrent with that, see if you can get a job working in the field. drop me an email and i can get you in touch with my company's recruiting/hr people (we qualified 5th at GC2 and we do a lot of UAS an UGV work). of the r&d staff, about 17% hold ph.d.s and over 90% hold at least one master's degree (there are several folks with multiple m.sc. degrees).
finally: good luck! i know i couldn't believe that i'm actually being paid to play with the toys and do cool stuff with them
AI might help with military simulations or game development, but real-world "robots" are mostly like overpriced remote control toys.
You should study real-time control (often done as a math or EE major) and embedded systems (often done as an EE major). Hmmm, how about getting an EE degree? Also, get to know Linux really well. Get to know VxWorks if you can find the opportunity.
Then there is the big thing you need to do: persistently apply for jobs in the field. Mostly you will be rejected with good reason, because you need experience to get experience. Keep trying.
You might settle for a somewhat less cool job at first. An example from the last time I was looking: some people had a Windows box as the user interface for a set of boards running Linux, which would do real-time control of a big device that filled orders for places like Netflix and Amazon. No, it doesn't walk or kill people, but it's a robot just the same. You need experience and a paycheck.
An autopilot is a robot. An anti-lock braking system is a robot. A safety interlock for an assembly line is a robot.
For better luck, try to be either a US citizen with TOP SECRET clearance (some military jobs provide it) or Japanese.
Go get yourself a degree in systems engineering. Pure AI research is within the realm of computer science and if that's what you want to do go study there. Robitics involve sensors and actuators, control systems, signal processing, embedded systems design, computer vision and image processing, and of course AI. None of those are outside the realm of engineering but pure theory is not our thing. This is after all an applied science and not a pure science.
----
Go canucks, habs, and sens!
First, there's two kinds of robotics out there. There's the classical robotics like factory automation systems, which are made to be as reliable and predictable as possible. If you want to work in that field, you probably should have gotten an MEng in mechanical engineering and gone on to a PhD within the field specializing in control systems.
Then there's adaptive robotics, closer to what people in general think of when they hear about robots. Adaptive, self-reliant systems perhaps capable of interaction with humans and able to get about in the world on the worlds' terms, not on their own. And there, I'm afraid, a CompSci degree is not going to cut it.
Doing real adaptive robotics is a hard problem. It requires a lot of different subspecialities, from mechanical engineering to learning theory to neuroscience. You may have physicists, linguists, mathematicians, behavioral psychologists and neurobiologists all on the same team. And most of them will know AI and programming well enough already. It would have been a lot easier for you had you focused on on a field like the ones above rather than CompSci, to be frank.
What you likely need is to be accepted at a PhD program with robotics somewhere; preferably one that lets - or requires - you widen your field to really learn one of the needed specialities as you go along. Since adaptive robotics is very experimental still, there really is no good training outside a PhD program, and nobody is going to take you seriously unless you have that doctorate anyhow.
Trust the Computer. The Computer is your friend.
1. enlist in the army as an intelligence analyst
2. get your TOP SECRET clearance
3. enjoy, um, visiting exotic places
4. quit the army
5. apply to Boeing, Ratheon, Lockheed, General Dynamics...
I would recommend getting an MS or PhD with a professor who already does robotics and publishes. The various alumni of CMU's Robot Lab would be good for this, as they have solid experience and lots of connections. (Most graduate students get their first job through their advisor, after all) And the larger, more established programs often keep a listing of their alumni. Since the newer ones are going to be trying to get tenure, they're going to be driven to publish. This can be good or bad, since it means that they'll be very enthusiastic... but your own interests may get swept aside.
Now, the actual degree is immaterial. Many of them are CS people, but just as many are electrical or computer engineers. There are some mechanical engineers too, and don't shy away from them -- they may need a good programmer for their current project.
Go visit these people in their labs, and ask about their current research, and tell them that you're interested in robotics as a career. Look at what they've published yesterday (Google Scholar is good for this) Talk to their grad students and see what they're working on.
You might want to consider locomotion biomechanics - it's robotics (the kind most people think of, not manufacturing) crossed with biology and is a lot more interesting imho. It's a very interdisciplinary field, and you have the opportunity to be exposed to, and focus on, biochemistry, evolutionary bio, forward & reverse kinematics (as someone mentioned), solving non-linear differential equations computationally (& efficiently), and building the circuitry, code and mechanical systems of a robot. Of course, it helps if you know some of this stuff going in - what you don't know, you'll learn from the diversity of people around you. For this, like any field, you will want to try to think of approaches to solve the problems facing the state-of-the-art right now and see if that excites you.
A (very) few biomech research labs (in the US: Berkeley, MIT, CMU, etc.) focus on robotics - contructing robots based on biological mechanisms and then trying to understand how to make robots better (or simply to understand biology better). Research projects involve some interesting issues, like replicating the action of muscles using man-made materials, mimicking the passive-dynamics of two-, four-, and many-legged animals, and even the mechanism of erectile tissue. ^.^;
While there are many good comments, robotics really isn't an industry. To design robots from the ground up is NOT easy. The Lego mindstorms setup will teach you basics. From there keep learning, start your own company. Despite what you know, there is AWAYS more to learn, always. This is truly an area where you have to make your own opportunities. The simple fact is there is no robotics industry, and that is an opportunity for you to help build it... Good luck
Support NYCountryLawyer RIAA vs People
Let me solve your question by solving a more generalized case. Let me reformulate the question more generally.
What should one do if one wants to prepare for a technical field that is in its infancy and changing so rapidly that any current technology is likely to be obsolete by the time one develops sufficient skills to create real applications?
Answer: study math. Math is the fundamental basis for all technical fields. Math skills never become obsolete, technical skills are obsolete the moment you learn them.
I'm a little surprised at the "There is no robotics industry" replies. I just graduated with an ECE degree, wanted to go into robotics, and had three awesome job offers from three awesome companies that are all really competitive in the field. I applied about ten different places in Boston alone that all catered to my interests, and I found many many more outside of the state. One big issue is location -- for robotics, in the US, you should be in Boston, Pittsburgh, or Silicon Valley. I'm in Boston, so I know the most about stuff going on here. Pittsburgh is building up a lot of spinoffs from CMU's Robotics Institute (much as nearly all the Boston robotics companies spun out of MIT). Silicon Valley has a lot of stuff coming out of Stanford. Then you have to decide what kind. Consumer robotics? Someplace like iRobot. Industrial robotics? Barrett Technologies makes ridiculous robotic arms. Places like Honda do pretty awesome things there, too. Military robots? Boston Dynamics makes the BigDog (it was covered on Slashdot a while ago). Places like Boeing have big contracts with the military's Future Combat Systems program. Draper Laboratory here in Cambridge does flying, swimming, and driving robots for military applications. Space robots? NASA, of course, as well as Ball Aerospace, Boeing, and various university labs. Medical robots? Vecna Technologies in Boston does a human-carrying bot for both battlefield and hospital use. Anthrotronix in Maryland makes robots for kids in physical therapy programs. Some company, I don't know who, makes the DaVinci surgical robot. I personally work on underwater robotics at a just-out-of-startup phase company here in Boston called Bluefin Robotics. Join the IEEE Robotics and Automation society to start networking. Google robotics conferences and see which companies attend. My delicious page here has a bunch of links to sites I used during my job search. I do only have my BS, but I plan to get an advanced degree later. As a few of my profs told me, if you want to do the really awesome stuff, or you ever want to start your own company (which I do), you need a PhD for the credibility. I also just feel like there are still other classes I want to take. ;) But MIT, CMU, Stanford, and Georgia Tech are all particularly well-known for their robotics and AI programs. Many other universities are starting to jump on that bandwagon as well -- BU is up and coming in surgical robotics.
So do your due diligence on Google, network as much as possible, and ask questions. There's a ton of stuff out there. Don't let anyone tell you the industry isn't there. This is just the beginning of something that will grow with incredible speed over the next couple of decades.
Ummm, sorry. Been a while since I commented on Slashdot. Forgot about doing HTML line breaks. ::looks guilty, slinks away::
I remember when the kids from the roomba company were posting group buys
on comp.robotics newsgroup, and look at 'em now. Yup, that is the right
answer. There aren't enough folks really trying robotics.
I worked in a car factory, at the time it was the most automated factory
in the US. I babysat the computer that sent work instructions out to the
robots. PLCs controled the robots, basically, they were programmed to
run certain stored programs and that was it*. The actual motion control
code was done by the robot manufacturer.
I've done AVG's and ROV's but that was either on my own, or at a technician
level. I am published, both in robotics journals, and in Servo magazine, it
doesn't really help.
If you have great ideas, and you want them brought to life do them on your
own. Build a UAV, or run the witbread unmanned or something really creative.
*funny true story. The guys doing the step programming for the rear window
installation had a few extra seconds after picking up the window. They
programmed the robot to pick up the glass in the wrong orientation, then
spin the glass an extra 180degrees, before installing it, more showy. It
was fun to watch. Well, one day there was some dust on the glass or something
and the suction cups couldn't grab the glass firmly, and it frisbee'd the
window about 20 feet across the factory. That weekend the robot got
re-programmed.
I have a small foundry for casting metal parts.
To cast small metal parts, you need to know pattern-making.
Patterns can be made from any material, wood being the most common.
If patterns are made from wood, then you need to know how to make wood patterns.
Wax, plaster, plastic, clay, polyform clay, found objects, etc.,
may be used as patterns.
Casting the pattern in metal? What about "draft"?
Numerous undercuts? How about flexible molds (RTV rubbers) or multi-piece rigid
molds?
Lost-wax casting can utilise technologies as old as 500 years or more,
or space-age technologies such as ceramic shell molding and casting.
If you can do robotics in software, you won't need to know how to create
parts for robots. I kinda think that robots are made from parts and that
the parts are driven by software.....
So, do you know anything about PCB's? Hydraulics? Five basic machines?
Logic gates? Hummmmm. Maybe some education in working with a group?
AI, CS, Robotics, Software, Mechanical Engineering, Electrical engineering,
group dynamics, ambition?
I just try to make art. Go figure.
You should find people doing work that you like and ask them. I don't know a huge amount about the field, but the requirements for a job with serious academic robotics researcher Rodney Brooks at MIT would be different than a job a his company iRobot, maker of the Roomba. And either one of those would probably be different than a job with Mark Tilden building toy robots, and different again from an industrial automation job.
Personally, I'd start with a visit to your nearest Robotics Society and have a chat with fellow enthusiasts about the shape of the field. Next I would start building my own robots to get an appreciation of the many facets of the work. Then I'd pick a few very specific skills, ones in demand with the kinds of teams you want to work with, and master them utterly. Deep skills give them a reason to hire you, and broad skills mean you'll integrate well in a team environment. And don't be shy about taking a job that seems a little beneath your skills if it gets you in the door with a group doing work you like.
Has AI and robotics advanced any further than what I read in the computer history books about MIT and Stanford? Last I read was that AI is/was still stuck in the same rut since the 1980s and the prefered programming language is LISP. Has the state of the art advanced in these two areas at all recently? Or is it still tuck away in obscure corners of academia?
If you plan on going into AI, you may as well skip the MS and go straight to a Ph. D. Lots of people don't realize they can do this, though it will probably put you at a disadvantage if you apply to very selective schools.
AI is very theoretical, and you will be better prepared for work in this area with a depth-oriented degree such as the Ph. D. You'll also get the feel of what it's like to do research in theoretical Computer Science, if you haven't already. Since many positions in AI involve research on one level or another, this is a good experience.
If you're more interested in robotics than AI, you may wish to consider an MS or MEng in Robotics, leading up to a Ph. D. in AI. This'll give you the engineering experience you need to build robots as well as the theoretical knowledge required to make pseudo-intelligent machines. It will probably come at the expense of more time - anywhere from 4 to 10 years is likely, so make sure you're prepared to devote that much time to your studies.
I think it depends on who you want to go work for - and that, in turn, can be influenced by your people skills. I think for a lot of white collar jobs, any halfway bright person can become a success if it's something they're interested enough to work at, so you already have that covered by making a goal of something you like. The real catch is, of course, getting hired.
If you consider yourself someone who can function well in a job interview, then IMHO the specifics of your credentials matter less. Aim for a smaller company, use your degree to get your foot in the door and then impress the person interviewing you with your dedication and flexibility. Larger companies, however, will have HR departments and resume screening processes - you won't even have a human laying eyes on your name if you don't have exactly the right degree, but if you DO, then the foot in the door is all you need - you'll be interviewed by an HR droid, who is not competent to judge your technical skills and is probably just trying to make sure you're not some kind of freak before rubber-stamping your way into the headcount.
Of course, these are generalizations. My point is, don't just envision what field you want to work in, try to picture what sort of environment (megacorp, 3-man shop, contractor, government?) you might thrive in. Being 'well prepared' means different things with different employers.
Perfectly Normal Industries
Build a giant killer robot. It may sound stupid, but the economics are totally there - people are willing to pay to avoid death. Right now though, you need to hit the books. Keep in mind that the robot needs to be able to withstand or avoid high energy military weapons: artilery, daisy cutters, lasers, and kinetic energy weapons. Don't worry about tactical nukes as long as you keep your robot on U.S. soil no president would dare use one: that would be political suicide. You will probably have to worry about corrosive chemical weapons: strong acids and oxidizers.
To have your robot withstand such destructive weapons you will have to study up on material science. Your robots exterior metals, plastics, and composites need to be top notch. You may want to create an cooling system for the electronic guts, but it may have to shut off and close vents when your robot is under attack. Stealth when not in populated areas, ordinance avoidance, and counter-measures should also be considered.
For AI your giant killer robot needs to be able to kill people, but it also needs to be able to collect protection money, and identify those who have paid protection money. Your urban navigation system needs to be able to avoid obstacles even while chasing down people who refuse to pay. Nothing says "don't pay me" like a giant toppled robot that can't get up. The most essential AI/mechanical problem may in fact be carrying customers to the nearest ATM machine without crushing them, but also without letting them escape.
Now for the economics. You really need to experimental with multiple business models once your robot is up and running - then see which is the most profitable. Maybe the most profits are in hanging around high income neighborhoods. Maybe a Paypal fund for an entire city - kind of like a telethon (once you reach $1 million no one in Des Moines will be killed). Remember though that once you hit a certain price people will be more likely to flee than pay - you may want to fund a phone survey to determine prices prior to operating in a certain location. Consider advertising - people unfamiliar with your product will flee with instinctual terror until they are aware of the commercial options.
So robotics as part of AI is indeed a very active area for research.
If you're talking about the most common (by modles and quantities) robots then you don't want AI as much as mechanical engineering or computer engineering.
If you're hopinh to get into AI with robotics you're probably looking at spending the next 10 years either as a student or researcher for a large university (CMU has been mentioned but others work too) and the degree you get should be based on having a well rounded knowledge of robotics (your PhD and research will establish a specialty).
If you'd like to do more consumer robotics you want computer engineering as it covers the relatively low level of mechanical design and high levels of electrical and software work that goes into consumer level robotics.
The last option is military or private company R&D in which case you'll need to specialize and work with a team of complimenting specialists so just chose to study in depth what it is that you like.
Now, before you do anything, remember that I'm 18 and haven't yet started on my B.S. in Computer Engineering. With that in mind, ask yourself why the above a) is wrong or b) wasn't obvious already.
-Tim Louden
I spent all of last year working on Robotics at my current company. I was asked to join the project because they needed someone with a good knowledge of linux.
On all the Robotics projects I was involved with, the first thing that was needed was to get the mechanics working. Most of my work during this phase was working with the mechanical engineers helping them to model the kinematics. Being able to talk with and understand mechanical engineers was important.
After the mechanical problems were solved, we had to provide the sensors, computer hardware and power that the system was going to need. Being able to speak the language of electrical engineers was important. I had to have a good understanding of embedded systems so that I could work with the electrical guys to determine the CPU and power requirements needed by the software.
From my experience, what I would recommend for anyone who wants to work in the Robotics field is learn as much as you can on your own, being able to express yourself to people who are currently working in the Robotics field is very important.
Get a lot of experience with embedded programming. Most of the software created for Robotics projects falls under the heading of embedded.
Choose a degree in Computer Engineering instead of Compueter Science. Usually there isn't a lot of difference between the two, except that Computer Engineering deals a little more with Electrical Engineering topics. I would also recommend a good background in multimedia communications and computer networking.
But the best advice that I can give is don't get hung up on the platform. Last year I was working on Robotics, this year it's software defined radios. Next year, if the funding returns, maybe I will go back to Robotics, who knows. In the past I've worked on telecommunications systems, aeronotics systems, guidance systems, infrared camera sysems, satellite communications systems and Robotics systems. Sure the Robotics was a lot of fun, but in the end, it was just another platform.
I'm hoping to finish up a Ph.D. in Robotics in the next couple of years, so I have some perspective. My background is C.S., and that serves me pretty well. There are not many interesting robots these days which are built and programmed from scratch by a single person. Robotics is by its nature cross disciplinary. You don't need to have the skills of a Mech E, or be a pro at VLSI work to be in the field. It's good to have some knowledge in those areas, but that's primarily to enable you to work effectively with other more hardware-oriented people, not to be a hardware designer. The other point of which you are probably aware is that Robotics/AI is an extraordinarily diverse field, even more so than CS, which itself contains a large variety of options. You'll be better off if you have some narrower focus for long term work, be that data mining, machine learning, biomimetics, or whatever else you want to do.
A comment that one of my lecturers made way back when I was doing my Cognitive Science degree always stuck with me. She said that what we consider A.I now won't be considered A.I in the future.
Take Chess as an example. Back in the 60s and 70s it was thought of as the cutting edge of A.I research. But when I was studying many moons ago it had long been realised that the ability to win a chess game did not help you solve other complex problems.
Self driving cars were considered science fiction ten years ago, are cutting edge now and will probably be passe ten years from now.
All I can say is don't get to locked into one vision of what robotics entails. Be flexible in what you learn and how you apply those skills to your job.
All the interesting positions I've seen in robotics require a PhD. I'm sure there are other ways to get in, but this is probably the easiest. It's not an easy field to get into at all.
Actually studying A.I. does not need a robot so badly. I normally use a combination of player and the provided simulators, stage and gazebo for A.I. http://playerstage.sf.net/ . Stage is good for studying group behaviour(agents) and gazebo is 3d with physics which make it great for more precise robotics.
For robot vision like tasks i suggest aquiring a fisheye camera, sticking it to some mobile platform - a chair with wheels in the worst case - for avoiding shake and wandering with it around, recording a movie to interpret it later with computer vision + A.I.
Also checkout my project http://miarn.sf.net/ to see my recent research in robotics, the world best GUI for mobile robots.
I am currently doing a thesis on building and controlling a bipedal robot. I was supposed to be getting my masters in mechatronics, but somehow it turned into EE controls. Very dissappointed because mechatronics sounds so much better. Anyway... depending upon where your robots focus is, it can contain Controls, AI, Dynamics, even Anatomy and Physiology if you choose that route. You should decide where your robot focus should be. Human-machine cybernetics? Human-machine interaction? Machine Intelligence? Sensory Systems? Locomotion? There are many different areas within robotics. It will take a while to learn them all if you choose to.
From experience I can say this: all of my training in EE, CS, and ME meant nothing until I tried to do something on my own. Thats the only way you can decide what you really like. Its also the only way you can really learn anything. So I say, before you choose a field to study in, put down the books and instead dabble a bit...
Funny!!!
this is great
Who says there are no jobs in robotics? That's hogwash. The industry is currently in the explosive growth phase, with Future Combat Systems, PackBots, Roombas, Scoobas and Pleos taking over. iRobot alone has about 40 job openings as we speak: http://robotstocknews.blogspot.com/2006/06/irobot- corporation-on-hiring-binge.html/ and many of their employees have become millionaires with the company's IPO.
-Thorn, publisher, Robot Stock News
...you don't need any fancy degrees or engineering jobs, the passion alone should drive you!
I've been a programmer for many years, and now (am lucky enough to) work in automation with big and powerful robots. I don't have any degree, but a passion for what I do, and people know that I'm good at it.
As much as I wish I could attend CMU or MIT and earn a PhD working with those brilliant minds, or be hired by Boston Dynamics or the like for one of their projects, I probably never will (because I have other passions, complete lack of discipline, and a wife and kid). But having been fascinated by robotics since I-dunno-when, I've always attempted my own research with ripped-apart radios and printers.
Now adays, I've committed a certain portion of my own income for my own robotics research. I've realized over the past year of designing and developing my own bot mechanically, electronically, and programmatically, that a PhD in each of these fields is definitely required! But I think I will learn more in my own research than studying in any one program. (Oh, and thank god for Google and sites like Stanford and CMU)
I guess my point is, maybe you won't ever get your dream job (though you will probably get close), but don't let that stop you from expressing your ideas and following your passions. Robotics is an exciting field with many fascinating facets, and the elite few robotics researchers and engineers in the world may not be able to think of everything!
Wow, a BS in CS with spec. in AI? Congrats, now go out there and compete with the Indians, Chinese, Vietnamese, various ex-iro curtain people, etc. Who get preferental treatment by America INC. because they can work for less than you. Sorry. Just from personal observation it would seem that robotics is at the point genetics was about 20 years ago. Kind of at the point where most of the work was experiemental and those working in the field were making peanuts. Like 20 years ago a PHD in Genetics would pull down a whopping 40k... That was before the Ludddites in the Pharmopoly finally read the memo that you can make drugs cheaper with microbes ( and Chinese slave labor ). Now if you can get the work, genetic work is usually pretty good, for now. As mor R+D is sold to the oversea slaves that may not hold up.
The mathematics you need to be able to do trajectory planning, joint kinematics, machine vision and all the other types of transformations are described in J.J.Craig's book (ISBN 0201095289). This is the application of stuff that has been known for fifty years. The interesting work is all in the application itself.
If you want something similar, but much more in demand, and scope to do something useful, you should look at plant automation - developing control systems for running industrial processes using PLCs. For example, the manufacture of cement consumes one percent of the world's energy, and yet it is controlled by incredibly backward technology. Process engineers that operate these systems are oblivious of all the developments in computing of the past thirty years. Their world is ripe for the injection of some of the expertise that has been developed in academia, and yet, the carry on with their PIDs, their SCADAs and their Ladder Logic.
Teach yourself a bit about grinders, mills, kilns and classifiers and do them a favour - bring them into the 21st century.
I recommend picking a problem and making a robot to solve it. Then SELL! SELL! SELL!
Robotics, renewable energy, innovative recyclable waste technologies are just what this country needs. These problems will likely be solved in time with innovative ideas that are sexy, neat, and astonishingly practical. Few companies are actively pursuing this type of goal. Those companies that are pursuing idealistic ventures are run by people who are basically part genius/nutcase/idealist/dreamer and are usually interested in hiring their like minded friends.
What this means to you is that the number of actual jobs in the area of robotics is small while the relatively large pool of applicants includes an assortment of real geniuses with great credentials. It does not sound to me like you can beat an MIT/Standford/Cornell/Harvard genius with great grades, great friends, and rich parents out of a robotics job.
I have a PhD from a prestigous school and know full well to pick my battles carefully. There is no way in hell I'd recommend you seek a PhD in anything remotely robotic unless you went to a top notch undergrad school, had great grades, great friends and rich parents. Many of those guys have trouble finding jobs; their advisors already have the corner on the available positions. Intrinsically interesting areas attract the best and the brightest. If you do not fit the bill the best way to make headway is to pick a problem, make a product and SELL! SELL! SELL!
My main academic interest was AI and I am currently working at a Japanese technology incubator, half for being able to speak English and half for that whole engineering degree thing. Specifically in AI I did a lot of work with natural languages. Those, um, don't lend themselves to working on the robots: of the 8 researchers I personally know on robotics the main commonality is a strong background in image processing (computational vision, etc) and 3D math (vision, motion, etc).
As to the availability of actual jobs in the industry: if you are bilingual and can program your way out of a paper bag Japan has a job for you. There are government incubators and private research labs (most associated with universities or laaaaaaaarge corporations) which have ooooodles of cash to throw at this field and not nearly enough qualified Japanese grad students to spend it on. If you can nail the language thing and get an introduction into the industry (which is not too difficult if you can handle the language thing and can program your way out of a paper bag), you're pretty much set.
As for pay and working conditions... well, put it this way, they vary wildly. I work ~35 hours a week and get paid about what a US liberal arts graduate would expect to make (not counting a compensation package offering decent perks like $50 a month housing & free insurance/taxes, so my actual quality of life corresponds to a salary somewhere in the $50-60k range, I'm guessing). I know another research at another institution who works a whole lot more for, well, a whole lot more. I'm sure there's also folks who work more for less.
Oh, incidentally, if you want something to burnish your resume a little bit and don't want to go into image processing human-computer interaction is pretty hot at the moment in my neck of the woods.
Help poke pirates in the eyepatch, arr.