Google, Intel, Microsoft Fund Robot Recipes
Dotnaught writes "Google, Intel, and Microsoft are funding what may become a robot invasion. Money from the three tech companies has enabled researchers at Carnegie Mellon University to create a new series of Internet-connected robots that almost anyone can build using off-the-shelf parts. These "recipes" describe how to build a robot that connects to the Internet using common parts and a $349 Qwerk controller from Charmed Labs."
...welcome our new internet-connected robotic overlords.
What is going to be the name of the first model?
Why on earth would I want a web connected robot in my house? Because I already have too much privacy? Because bored kids would never think to trash my house with my own robot? Why not just install webcams and tape a web controlled taser to my neck..
We are all just people.
brings new meaning to "bots" doesn't it.
Sigs are too short to say anything truly profound so read the above post instead.
$129 - Nintendo DS
$54 - Supercard-Lite-MicroSD
$15 - 1GB microSD
$49 - DSerial2
$99 - RoboDS
---
$350
Combine with the open source full linux wifi environment, and I don't know why you'd want to spend $350 on that controller (I'm lazy and haven't even read the specs on the thing, but seriously, I can't imagine there is anything the roboDS can't do that it could)
http://www.natrium42.com/shop/robods.php
-dmc/jdog
Do you want your house trashed
Accept Decline
Engineering is the art of compromise.
what about itsatrap?
I know they have robotic studio out now but...the Querk is Linux based. wtf are they thinking?
I think MS is just dishing out some cash to get in on the action but other wise
has nothing to do with this. It's just another promo stunt for them so they can proclaim
they play 'open' just like they did with robotics studio. In reality they had no
chance of winning in that arena because play/stage/gazebo was already out and is
better built for people who do serious robotics. Or this is their big promo stunt to
now start charging joe shmoe for the robotics studio..... either way it still stinks.
Money is the root of all evil?
You can do basic motor control with a $10 microcontroller. Add a stepper driver IC and you might be up to $25-30.
Basic on/off DC motor control is a $1.50 transistor away.. and hell, why not use the LinuxCNC project if you want to do really complicated control?
I'm not exactly sure what MS is trying to do here. Robots are cool, and a great interactive toy - but you waste a lot of resources building them if you want to explore AI and the like - it's easier to simulate concepts.
..don't panic
That should be the title of this piece of news.
if I could only find a cheap supply of red LEDs for the eyes.
Recycle PCs and build a wireless community network www.hillsborough.org.nz
Whilst it's laudable that companies are investing in robotics at all, it seems to me that the time has come for investment on a commercial scale in robotics for specific applications. These 'hobby' type robots are all well and good (and no doubt particularly appealing to many around here) but they don't actually DO very much of any use, and the average member of the public is not going to be all that excited by them.
Roomva and similar robots are a step in the right direction, IMHO: relatively cheap, one- or two-function robots which have an obvious and straightforward function. People can see that, understand it, and if it works well (which I gather is not really the case just yet), will want to buy it. Once there's actual profit to be had, investment will increase rapidly and voila, the real robot revolution* begins.
We seem to be at a point where we have the tech for some truly cool everyday use robots. Perhaps even something like an x-prize for robotics, with the objective being to build a cheap, mass-produce-able, functional robot to perform a specific household task, would do the trick. Some major investment from some major players could kick start a very fundamental change to the way we live.
Plus, having lots of robots around the house would be frickin cool...
* the good kind, not the humanity-crushing kind
Read Pynchon.
Robot Universal Remote (RUR) == Rossums Universal Robots
Money is the root of all evil?
As someone marginally involved with the project (my wife is Illah's student), I thought I might clear up some confusion that I've seen in the comments so far.
First of all, the writeup focuses a lot on the funding from Microsoft, Google, and Intel, but in reality these robots' hardware and software are being designed at Carnegie Mellon. Specifically, Microsoft isn't involved with any of the programming and design. The robots indeed run Linux, not Windows. They're not using the MS robotics studio to develop the robot.
This robot is primarily geared toward education: interesting kids in science and robotics. The project lead (Illah Nourbakhsh) has extensive experience creating science museum exhibits, tour-guide robots, and other forms of human-robot interaction. The networking functionality allows users to teleoperate the robot easily from a web browser.
Right now, the project has obtained funding for curriculum development, integrating the TeRK into beginning computer science courses at the junior college and university levels, in order to increase the appeal of computer science to people who might not otherwise be interested in the field.
-- I'm not a pessimist, I'm a realist. It's not my fault that life sucks so much. --
I, for one, welcome our new Microsoft Google Robot BETA with Intel inside overlords.
Well, there is a BSOD screensaver for linux, and we could define an init state with no processes. So we could fake a BSOD after turning on the gas in your home.
When our name is on the back of your car, we're behind you all the way!
Google, Intel, And Microsoft Fund Robot 'Recipes'
j html?articleID=199201449
Abstract: Money from the three companies has enabled researchers at Carnegie Mellon University to create a series of Internet-connected robots that almost anyone can build using off-the-shelf parts.
By Thomas Claburn, InformationWeek
April 25, 2007
URL: http://www.informationweek.com/story/showArticle.
Google, Intel, and Microsoft are funding what may become a robot invasion. Money from the three tech companies has enabled researchers at Carnegie Mellon University to create a new series of Internet-connected robots that almost anyone can build using off-the-shelf parts.
As part of the Telepresence Robot Kit (TeRK), a joint effort unveiled last summer between the Carnegie Mellon Robotics Institute and Charmed Labs, associate professor of robotics Illah Nourbakhsh and members of his Community Robotics, Education, and Technology Empowerment (CREATE) Lab have created a series of "recipes" for robot building. (Those who recall The Twilight Zone will be relieved to find that "To Serve Man" is not among them.)
Possible robots range from a three-wheeled model with a mounted camera to a sensor-equipped flower.
The project's goal is to expand involvement in robotics.
The heart of the TeRK is the robot controller, called Qwerk, available from the Charmed Labs Web site ($349). The unit functions as an electronic brain and handles wireless Internet connectivity, motion control, and functions like sending and receiving photos or video, responding to RSS feeds, and searching the Net.
Qwerk is a Linux-based computer. It uses a field-programmable gate array to control motors, servos, cameras, amplifiers, and other devices. It also can accept USB peripheral devices, such as Web cameras and GPS receivers.
"We leveraged several low-cost, yet high-performance components that were originally developed for the consumer electronics industry when we designed Qwerk," said Rich LeGrand, president of Charmed Labs, in a statement. "The result is a cost-effective robot controller with impressive capabilities."
The robots are intended for practical uses, in addition to education and entertainment. They can be used for home or pet monitoring, for example. A future recipe being developed includes environmental sensors for measuring noise and air pollution.
Nourbakhsh doesn't subscribe "to geeky notions of what robots should be." That may explain one of the recipes that he and his team are working on: a controllable stuffed teddy bear.
Be afraid.
when the robot crashes, it emits the Blue Scream of Death.
If you think the "browser wars" were bad, just wait untill Microsoft and $POTENTIAL_COMPETITION fight over control of a household robot.
try { do() || do_not(); } catch (JediException err) { yoda(err); }
I'm always lookin' for a new way to eat robot.
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
I think I saw this in a movie starring a governor. They called it AeroTubes or something like that.
Unfornately this gives us a whole new definition of
Spamming Net Bots
hey everyone, its a story on internet connected robots... lets talk about how shitty mac's are!
there must be some way outta here, said the joker to the thief
I contributed to TeRK while working on my MS at CMU.
The idea is to provide as simple of an interface to programming the robot as possible. You can write your own stuff directly on the hardware if you like (it's got a serial connection so it's easy to connect to). Or, you can take advantage of the layers of code and write something which runs on your PC... but still has access to things like values from the analog inputs and moving the motors -- all via 802.11. The project uses a lot of open source and the source code for all of the components is available. There is a lot of framework code written in C that runs on the Qwerk board itself, and it uses ICE to connect from the board to either a relay server or your PC. Then, for the people who don't like to program at all (or are just starting out), there is a lot of software, including a basic emulator of the board, mostly written in Java, that they can just run on Windows, Mac OS, or Linux.
During development, we took our PC app and a couple of Qwerks to a group of robotics hobbyists and they were floored by the kind of capability you can get for free with the Qwerk and all of the software that's already been written. Most of them wanted to find a way to incorporate the board into their own projects.
Anyway, the goal of the project is to have a wide appeal. I hope it can get a lot more people excited about what they can do, and all at a very low cost compared to other kits.
// No comment
Are there any specs for the roboDS? I don't see them anywhere, not to mention it isn't shipping yet.
o ntent&task=view&id=29
Qwerk info here:
http://www.charmedlabs.com/index.php?option=com_c
In short, 200 MHz ARM9, 32 MB RAM, 16 servo ports, 4 motor ports, 16 digital inputs, 8 analog inputs. And it's actually supported for this purpose instead of relying on Nintendo putting up with your hacking.
But, in the DS's defense, having a builtin screen *is* kind of cute.
This configuration is what I have in mind for a modern small scale robot.(Small scale because it doesn't have large memory storage on board ) Unfortunately this kind of robot is totally useless without a huge database , server and wireless connection. Robot with this hardware configuration IS a robot!!! I go crazy when I see people wasting time with something like 2 servo motors , small cpu like 8051 with 2k of memory and few sensors that help em not to bump into a first obstacle. Please use your imagination and time for something better.There is still lot's of stupid coffee, washing machines out there to make em smarter.I just don't understand why most of the people want to create such a stupid robot first thing when they learn how to program MCU's.Anyway I hope database and software will be open for others to hook up with it in the future so this robot hungry hobbyists can create something useful. Check http://www.fira.net/. Those robots are connected to central computer with wireless connection and play soccer.
Carl and Tom spoke at one of the Pittsburgh dorkbot meetings. There's iPod video and MPEG-4 of their presentation in the March archives.
Zhrodague.net - I do projects and stuff too.
MS is "sponsoring" with software licenses! Windows and Visual (robotic) Studio isn't real money!
Back in 1996, I tried talking a couple of friends into setting up a site to sell DVDs. My argument was that DVDs were going to take over the VHS, and ppl were going to do libraries of these. By selling on the net, we could sell more for less (individually). But the big argument that I made was that with these being standardized, that after the first year, we could robotize the fulfillment system. That is, once the order is placed for say 10 DVDs, a robot system would extract from racks the appropriate dvds and then send it to be packaged. I was even thinking of multiple size box es for 4 or larger would accomplish it, this nicely. This would have been very useful in that case.
I prefer the "u" in honour as it seems to be missing these days.
Get that Old Glory Insurance now.
I'd think that in the long run, it would be more useful for the robots to contact each other remotely than to be web-controlled by a human. And don't worry, Judgement Day already didn't happen.
09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
But a robotic chair? Can we not even sit down without the help of a technology? I guess the point is that they roll back to a corner out of the way for the cleaning maids after hours... but what about when a meeting goes overtime and while everyone's sitting down suddenly we all start shifting towards the wall? Even if there's safeguards for this, there's sure to be some failures... Someone save the office space from the coming invasion!
Reading the specs, the Qwerk is a fairly well endowed Linux-based single-board computer with lots of I/O capability, plus an FPGA.
Since the FPGA is the only thing that really separates it from a huge number of other small form-factor industrial SBCs out there, what exactly does the FPGA do in this application?
You hear a little voice come out of the Googlerobot: "This robot has violated the First Law, you should ask Google why they put out such a crappy product."
And the rest is (future) history.
"The most sensible request of government we make is not, "Do something!" But "Quit it!"
Hey I'm going build a wifi controller for this so I won't need a computer; It will have a few sticks one that makes the left wheels turn one the right. Then I can control it from my couch I'm going to call it the Remote Controlled Car. I like the idea here but I'd have to say it's not a robot until it can process some form of independent thinking. From what it sounds like it just takes my commands and sends back video. I have to control it so if I want it to go find my book I still have to drive it around the house and identify watch a video as I identify my book. When it can learn to drive around the house on its own (like that robo vacuum) it will be a robot when, it can drive around the house and pick out object I'll buy it. But right now the only thing to see here is a nice interface which hopefully can be modified to control any number of devices I can plug a wifi connector into. Good first step but it needs a brain and something practical to be great. When we have a open source program that lets it avoid obstacles on it's own I'll be more interested. It needs to beat out the Lego Mindstorm.
Eh, the "other half" of CU robotics--utter crap. What a joke. CU has some really cool shit, and you're listing this? Their soccer stuff is more evolved than this Qwerk hardware. Hell, 3 year old motherboards with common motor/stepper/servo interface boards can do more than this Qwerk board.
re the DS
It's more like $250-- You can maybe drop the $99 RoboDS part, since that's actually parts for the robot (the $49 part is the interfacing part, the $99 kit includes servos, and the like) although it does not include power, while the Qwerk does. The other robot parts included with the $99 are not included in the $350 CharmedLabs Qwerk solution.
[Although I'm at a loss how WiFi on the DS interfaced--last I saw, that was still being worked on, but I guess it's been solved (if someone knows a direct link, let me know, since I'm lazy).]
The Qwerk, besides having the power interfaced, is a more capable platform all around...4 motor controllers, better interfacing, etc. It would be nice if it had more motor controllers. It's nice, but just which manufacturers had daughtboard add on for your choice of motor interface instead of trying to interface a limited or obscure format that no one really wants (RC servos are fine for many things, but not for a LOT of robotics; in turn, 4 motor controllers is way too few).
Overall, it seems it would be easier to pick up an older EPIA board and a stepper controller and interface everything from there. Would take more work and be a slightly larger hardware package in the end; going to a Nano would be fine but blow the budget.
What is the problem there?
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
* Vacuum (we sort of have that but I have dogs so i need a real sized canister)
* Clean and stack Dirty Dishes
* Fold and hang clothes (I can wash and dry them).
* Mow lawn (we sort of have that)
I can't see buying a robot for fun. But I would pay about $300 to $500 per item on that list.
She was like chocolate when she drank... semi-sweet at first and then increasingly bitter.
a link would have completed it.
You better watch out, there may be dogs about . .
I'll admit, that the development environment for the NDS is lacking, and that would be a big selling point for the Qwerk at this moment. But what is important to keep in mind is that the NDS dev env isn't lacking due to any fundamental DRM or otherwise locked down or obfuscation problem. It really just needs some more elbow grease from the community to polish what we already have. I'm personally doing my small part to that end.
Now then, to answer your question about the RoboDS which isn't _quite_ shipping yet (This person/people have a good track record however, the DSerial(2) had a similar pre-order/ship-date, and they absolutely hit it. Also, the schematic and parts list for the DSerial are available, so there is no reason that were it popular, any manufacturer or university, couldn't produce however many they needed to fill demand.
As for specs, it's probably best to look at the specs of the NDS and DSerial2. The RoboDS is just a few rudimentary physical parts. Abbreviating the NDS- An ARM7 AND and ARM9, 802.11, 2 LCD screens, one with touchpad, microphone, stereo speakers, headphone jack, buttons.
The Supercard-Lite-MicroSD gives you up to 2GB(+?) flash memory support, 32MB of ram (in addition to the 4MB that the NDS has natively)
And cutting and pasting from the natrium website regarding the DSerial2
# 8051 microcontroller running at 24MHz
# Reprogrammable from DS, firmwares available at NaWiki
# Free development tools available
# 18 GPIO lines, 2 status LEDs
# UART with RS-232 level converter (can be disabled)
# Full-speed USB 2.0 device
# PWM and ADC available
# 2D tilt sensor
-dmc/jdog
Here is the not tremendously mature code for doing wifi on the NDS.
On the other hand, there are a couple articles in the slashdot submission queue pointing out the release of homebrew quake (with wifi) for the NDS. So its fair to say it's a solved problem. But again, there is a fair amount of work to be done as far as releasing a truly polished, complete, easy to use development environment for the NDS. What is there is certainly not bad, but it is a bit rough around the edges.
http://akkit.org/dswifi/