Eddie: Hope or Hype?
I am still waiting to form my opinion on Eddie. I wish Parallax & Microsoft all the best on this one, but honestly I don't understand the mad rush to get one. It hasn't demonstrated any amazing abilities yet, there is no "killer app", no arm, no Northstar navigation AFAIK. Sure there's a contest and a flurry of activity and all the early adopters want to be the first on their block. The last time I saw this much anticipation was for the 914 pc bot, which overpromised and underdelivered on all counts. Their marketing campaign said "what will YOU do with it", or "the only limit is your imagination", and it meant they didn't know what to do either. Lots of disenchanted 914 owners out there. BTW, the 914 has changed hands at least twice since Tom Burick launched it. It was badged Whitebox for a while, and now it's flying the Heathkit HeRobot colors.
Evolution Robotics ER-1 was really the first laptop on wheels that really didn't get anywhere, despite Northstar navigation and a bunch of smart & enthusiastic early adopters.
Other than Kinect and Microsoft Robotics Studio (on which the jury's still out), what's changed since those two (over)promising robots were orphaned? Am I missing something?
Evolution Robotics ER-1 was really the first laptop on wheels that really didn't get anywhere, despite Northstar navigation and a bunch of smart & enthusiastic early adopters.
Other than Kinect and Microsoft Robotics Studio (on which the jury's still out), what's changed since those two (over)promising robots were orphaned? Am I missing something?
Comments
Erco,
Eddie may be the reference platform, but it is not the only robot platform that conforms to the reference requirements. MSRS, because it is free, will probably be accepted more widely than other robot languages that an experimenter must pay for. However, I believe that Eddie is a prime candidate for Robot Basic too where the Eddie Propeller Controller Board is commanded from a more powerful remote computer. The only problem I see at the moment is how to remotely converse with the Kinect which might be done by a wireless link maybe?
Just a thought.
Regards,
TCIII
http://www.mobilerobots.com/ResearchRobots/ResearchRobots/PioneerP3DX.aspx
which is very popular in the educational and industrial research markets.
I don't think you can compare against the Whitebox products, which have become very expensive, and the ER-1, which in its day was quite popular. They sold a lot of units, even without Northstar. Evolution has instead concentrated on efforts -- mostly software licensing -- that provide higher profits. You might be surprised how little margin there is in selling amateur and educational robots.
I doubt Microsoft wants to get into the robotics hardware biz. They'd be delighted, though, if their OS were used in as many machine applications as possible. I make the point in the introduction to RBB4 that building robots has become less about the actual robots, and more about the technologies involved in them, as those technologies can be applied across many disciplines. Building bots is just a fun way to play with gadgets.
-- Gordon
At least 200 times the number of hobbyists who can afford $400k for a PR2 from Willow Garage!
http://spectrum.ieee.org/automaton/robotics/robotics-software/willow-garage-sells-first-pr2-robots
All we need now is an Eddie one that can generate enough revenue to login, buy and army of parts, and assemble it all.
Of course a good doctor would be useful too.
Thanks Parallax for this great project.
I feel that Eddie's basic software, and op system has taken the hobbyist's robot to the level of the latest computers, where robots are becoming useful tools in business and households, where the function of the robot is more productive, than the invention of its wheels.
Yes, I have to be older than sin to have built IMSI computers.
Robert
Did you notice that article was written by ERiCO?
I am right up there with you. I had a MOS Technology KIM6502 that I had converted to a KIMSI with an interface that connected the KIM6502 to an IMSI back plane. With the KIMSI is was able to take advantage of most of the memory boards and I/O cards that plugged into the IMSI bus. I even had a single sided 5" floppy attached through a Western Digital IMSI board.
Regards,
TCIII
No, but that heap big funny stuff, DD!
I'm more interested in making consumer robot products for the mass market, where I'm convinced KISS will rule forever. Everyone likes robots, but not everyone wants to be a roboticist. iRobot's Colin Angle told me his reasoning for why Roomba is so popular: simplicity. You push one button and walk away. They have done numerous tests that prove that high tech (& pricey) room mapping does not significantly improve cleaning over Roombas bump switches & simple onboard algorithms. I know better than to argue with a successful millionaire CEO from MIT.
Speaking of which, where's Helen these days? http://www.xconomy.com/boston/2009/02/09/irobot-co-founder-greiner-launches-stealth-robotics-company-the-droid-works/
No more than the Hero, which predominately sold to schools. As a hobbyist you have to be pretty well off to afford the $1,000 for the Eddie, $150 for the Kinect, whatever for the laptop, and $50-500 for batteries and various other sensors and do-dads you want to add.
Like I said above, this is a platform for education and research. Hobbyists aren't shut out, of course, but there are far cheaper alternatives that can provide the basics.
Eddie (and platforms like it) have the potential to help develop applications for consumer robotics, though I think its feature list is more appropriate for higher-end products than an automated rug vacuum. The Kinect opens a lot of potential for robots that before required extraordinarily expensive sensors. The Kinect can determine the distance, direction, and motions of a golf player, for example, and could make for a fairly robust automated caddy. They already have these, but they're expensive and not always reliable.
Vision is the killer app for a robot, and essentially Eddie is a mobile platform for experimenting with the possibilities. And a fairly low-cost one. Like I said, it's about a third of the price of the most popular research platform.
-- Gordon
I don't know if it's very clear from the videos and information you've seen about RDS, but they're into robotics for the long-term. Their focus is on problem-solving and utility, and you can see that it's going to take a while to achieve results like carrying mail around an office without lines, magnets or beacons. I think the Microsoft solution has much in common with the Japanese focus about caring for people, too.
PC-based robotics isn't my interest and probably not yours, but for some problems I've seen presentations from Microsoft that help me understand and accept their vision. They have 115 engineers in their robotics division, so there's some serious momentum to make RDS a success.
As for the actual Eddie reference platform, we don't get into many speculative business arrangements given the importance of our microcontroller business. So far we are receiving the positive results we expected from a sales standpoint.
Ken Gracey
Prop boe I think will be my next platform.
Where are people's sense of adventure?! Microsoft would vastly prefer you connect a Kinect to Windows, but it's not like you NEED Windows to interface to it. There are several hacks for it for other OS's, including Mac and Linux.
So it says Eddie is for Windows running RDS. When did that ever stop anyone from being brave?
-- Gordon
Runs on Windows.
Just a thought.
Regards,
TCIII