Slashdot Mirror

The cheapest 2D LIDAR you're going to find is about $2000 (Hokuyo URG). It has pretty terrible range for a LIDAR, but it's still a good sensor. For 3D ranging you're going to spend at least a grand. The IFM O3D 2XX is the cheapest 3D Flash LIDAR I know of, and you're only getting 64x48 pixels of resolution, so essentially 1% of the pixels you're getting back from the Kinect for 10x the price. Given this, the Kinect is truely an amazing sensor.

I am not doubting that the Kinect is changing the game. I just question whether it is revolutionary or not.

Kinect is changing the game but it's not revolutionary? What's your definition of a revolution then? Before the kinect, it cost me close to $10,000 for a good 3d point cloud data. If I had more room on my robot, I might put a Hokuyo LIDAR on a pivot but that still put me back 6 grand. Today I use industrial sensors from IFM, re-purposed for Robotics. They cost about $1500, and only provide 50x64 pixels of range data, as compared to the Kinect's 320x240.

So the cheapest feasible sensor I can buy costs $1500. So here comes Microsoft. They're selling a sensors 10 times cheaper with 24 time the resolution. Now any old schmuck can buy this and test their idea for a new image segmentation algorithm. This has NEVER been possible before.

So yeah, Kinect is changing the game. That's the definition of a revolution. Just because it was done in a lab before by Ph.D.s after 10s of thousands of dollars of time, effort, and equipment doesn't diminish it. If a company started selling robot cars to the public, that would be revolutionary too, even though we can do that in the lab (for $1,000,000+).

And Microsoft can't get all the credit; none of this would be possible without ROS and the amazing Point Cloud Library. This is a second component of the kinect revolution, which, in itself is revolutionary.

I am not doubting that the Kinect is changing the game. I just question whether it is revolutionary or not.

Kinect is changing the game but it's not revolutionary? What's your definition of a revolution then? Before the kinect, it cost me close to $10,000 for a good 3d point cloud data. If I had more room on my robot, I might put a Hokuyo LIDAR on a pivot but that still put me back 6 grand. Today I use industrial sensors from IFM, re-purposed for Robotics. They cost about $1500, and only provide 50x64 pixels of range data, as compared to the Kinect's 320x240.

So the cheapest feasible sensor I can buy costs $1500. So here comes Microsoft. They're selling a sensors 10 times cheaper with 24 time the resolution. Now any old schmuck can buy this and test their idea for a new image segmentation algorithm. This has NEVER been possible before.

So yeah, Kinect is changing the game. That's the definition of a revolution. Just because it was done in a lab before by Ph.D.s after 10s of thousands of dollars of time, effort, and equipment doesn't diminish it. If a company started selling robot cars to the public, that would be revolutionary too, even though we can do that in the lab (for $1,000,000+).

And Microsoft can't get all the credit; none of this would be possible without ROS and the amazing Point Cloud Library. This is a second component of the kinect revolution, which, in itself is revolutionary.

From a technical perspective, this is just such a pathetic response from sony to the motion controller game.

Sony used to be such an innovator (or so everyone tells me) but all I have seen from them for years is pisspoor effort after pisspoor effort. This is a particularly sad effort on their part.
Why do I say that?

From a motion tracking point of view, tracking a brightly colored ball is pretty much the simplest possible thing you can do. Check out this embedded system you could buy for $150 or so in 2002 that did it as a basic demo:
http://www.acroname.com/examples/10067/10067.html

It was one of the first things I did when I learned how to use the OpenCV computer vision library. Its just pathetically easy to do. You basically max out the contrast, and any pixels still white are the bright spots. Go check out ball tracking or blob tracking videos on YouTube. Every college student with a class in MATLAB has probably learned how to do it from the ground up, without a library.

Meanwhile, Microsoft is creating Kinect, which combines multiple cameras to create depth and color maps of your living room and model your entire skeleton in real time. *That* is incredibly complex and extremely innovating!

It is just so sad that Sony is actually releasing this as a product. It is literally like someone said "Hey, we need to do something about the wii", and someone said, "okay, how can we do motion tracking the cheapest way?"

In fact, that's probably why they did it. Instead of putting an infrared tracking camera in each remote (like the Wii), they can just use one camera on the TV and just put LEDs in the remote. They probably did this first and foremost because it was cheap, and for no other reason. Its sad that a company that used to innovate is not just a cost-cutting me-too company. They didn't think about how to improve on the concept, or if it even made sense; they just copied it with the least cost they could.
-Taylor

In my humble opinion, the Darpa Grand Challenge, by offering a market to LIDAR makers, made vision-based SLAM a thing of the past and the under-budgeted.

That's what many of us with Grand Challenge entries once thought. Even Sebastian Thrun once thought that. But, in fact, the winning 2005 Stanford "Stanley" vehicle was running mostly on vision. Above 25MPH it was out-driving its LIDAR range. The vision system wasn't doing SLAM, though. It was comparing the road further ahead with the near road. If they "looked the same" (the machine learning system for making that judgment was the breakthrough) and the LIDARs profiled the near road as flat, then the vehicle could drive faster than it could stop within the LIDAR range.

For the Urban Challenge, LIDAR units were more useful, because the speeds were slower and the environment more cluttered. But see the current issue of IEEE Trans. on Robotics, the special issue on SLAM, to see how much progress has been made. It's useful to use a camera and a limited LIDAR together with a SLAM algorithm; the vision system brings in more data and the LIDAR has less ambiguity.

The Velodyne thing (which is a better-built version of the Team Dad spinning LIDARs of 2005) is a good device, but too big, too expensive, and has too much rotating machinery for a production product. I've met its designers and seen the thing. The next step will probably involve either flash LIDAR or MEMS mirrors. Eye-safe flash LIDAR is a reality, and if produced in volume, it wouldn't be that expensive. It's expensive now only because it needs custom ICs.

An affordable little non-scanning 3D LIDAR for indoor use would be useful. There's the Swiss Ranger, the first device that qualifies. This is a true 3D time of flight sensor with no moving parts and 176x144 pixels. It's been around for about five years as a custom research item, but it's now being sold as a product by Acroname for $7500. The price needs to drop by an order of magnitude or two, which is quite possible.

I love the Brainstem, it's got 5 digital I/O ports, 5 analog I/O ports, 4 servo outs, GP2D02 port, RS232 port and an IIC bus. All of the functionality you could need, such as sonar, IR, servo, text-to-speech, even pyroelectric (human heat-signature sensor), and many more are implemented in pre-built, extensive libraries. You can run 4 VMs at once and swap them out among 11 different 1MB programs, with 1MB of shared scratch space. You can also program low level "reflexes" to run independently of the VMs. They're programmed with the TEA, or Tiny Embedded Application, language (you "steep" .tea files into .cup files, which are the binaries :-D), which is a stripped-down, easy to use, C-syntax language that's really easy to learn and program with. The compiler is just a tiny little console that can be downloaded for free and even runs on PalmOS. Did I mention they're really easy to interface with Palm Pilots? They're great for post-intro robotics classes when you want more functionality and would like to really get down into nuts and bolts construction. As well as being cheap and easy to use, there are a ton of kits and libraries for platforms that specifically use the Brainstem. After doing research at my university's robotics lab with several different platforms for over 3 years now, I'd recommend these over anything else, of course, that's just my personal opinion and there's a ton of other great controllers out there, especially the BASIC, JAVA and other Stamp controllers. Here's the link http://www.acroname.com/robotics/parts/S1-GP-BRD.h tml -Julius

I know this is a little bit off subject, but in the interest of DIY you could build your own. I've been using Sharp's OEM IR sensors to good effect in my indoor lighting projects. They're basically a frequency modulated IR beam emitted from the sensor, coupled with an IR sensor configured to detect only the frequency of the emitter. The devices output a voltage which is proportional to the distance detected. If you know how to program a microcontroller, you could program your own object detection algorithm which would likely trump most of the commercial devices out there.

Anyway, my two cents. I'm using the shorter range versions of the sensor for my lighting design (http://www.designeb.com/).

More information on the Sharp distance sensors can be found here:

http://www.acroname.com/robotics/info/articles/sha rp/sharp.html

Imagine what it'd do for the economy if they reinvented the wheel!

Instead of using passive IR sensors... use (gasp) active IR sensors or ultrasonic - like autodoors today use.

The door in the article appears to be using (active) Sharp IR distance sensors. I wonder if some of their problems are the large number of sensors interfering somewhat.

I'm assuming it's referring to an infrared range sensor, which operates like a simple variant of radar.

Lego Mindstorms can be a great way to start. But it can be limiting. It comes with only 2 motors. It can control only 3 motors and have 3 sensors. The programming environment can be limiting too, but there are other options on the web to help.

If you think your friend would be interested in something a little more complex, I can point you to some other resources...

There are some robot kits on the market. A pretty good one is: SumoBot Robot ~$150

Or you could get a platform and controller board seperately:
Tracked Motorized Platform ~$60
Controllers The OOPIC II Starter package ~$70 is a decient starter

A number of hobby clubs in Colorado like the Atmel Atmega Microcontrollers for building custom robot board such as the DARC Board.

The Portland Robot Group has a lot of great information for people starting out with hobby robotics.

My robot blog with lots of pictures.

Lego Mindstorms is a neat way to start, but it can be a very limiting platform. For the most part you are limited to 3 sensors and 3 motors - and the base kit only comes with 2 motors.

If your friend would be happy with Lego robots, that would be a good choice. If you want to look at some kits that allow a lot more interation on more of the complexities, check out:
Parallax - SumoBot Robot ~$150

Also you could look at mixing and matching:
Budget Robotics - Tracked platform
Acroname - Controllers

There are other options, especially if you friend would be interested it doing more work from scratch.

-- Keith - My: Robot Blog Main Blog

http://www.oopic.com/

http://www.acroname.com/robotics/parts/S1-GP-BRD.h tml

These are two microcontrollers for robotics. I have used the OOPIC and heard that the Brainstem is also very good. Both come with sample code. They are an excellent way to start a robotics project.

Try the following places also, Acroname and Mondotronics Robot Store

David Culp

Coach of the 2004 Oklahoma Regional Botball (http://www.botball.org) Champions (1st and 2nd place teams actually).

The problem with the watches is that they are analog. You can't see them in the dark, and they don't have alarms. Check out the Mars alarm clock.

Sounds a lot like the products for sale at Acroname in Boulder CO.
Specifically, the Palm Pilot Robot Kit, or PPRK

Never played, but they look like fun
--

Re: Same Robot Used to Search WTC site

Dammit, I forgot to include two links:

Robo02: http://www.robo02.com/

Acroname: http://www.acroname.com/

Sorry about that...

Ok, it hasn't been commercialized but it isn't as whimpy - check this out "Vision is hard," Gross said. "Nobody has succeeded in making it work in real time." One word: Aibo Oh, and give CMU some credit Jeff

Jeff

Acroname makes some interesting robotic stuff. Here's a palm-pilot robot kit that they produce.

If you really want to know how to build autonomous robots, there are several competitions with the same level of excitement as Battlebots et al, but with the added benefit that you can imagine it might eventually be a good thing to turn the resulting machines loose in the real world.

For instance, Robot Sumo is quite popular in Japan and the US. You can find the rules and links to competitions at Sine Robotics. Another big competition is the Trinity College Fire-Fighting Home Robot Contest, wherein robots navigate a known maze (a model house floorplan) to put out a fire (simulated by a lit candle).

Another nice thing about these competitions is they can be cheaper to get into than Battlebots (less heavy iron and welding). And there are lots of good people and organizations who can help you climb the learning curve. Just a few of my favorites are the Seattle Robotics Society, The Robotics Club of Yahoo, Raleigh Triangle Amateur Robotics Group, Portland Area Robotics Society, Robotics Society of Southern CA, and the San Francisco Robotics Society of America.

Finally, here's a few places you can find parts, books, plans, kits, and lots of links: Mondo-Tronics, Acroname, and Robot Books.com.