An awesome, futuristic, all-in-one robot chassis

bot

No matter how many advances in electronics we find, we’re constantly surprised at the lack of progress in robot chassis. Sure, it should be a simple task to make a capable robot that looks cool, but aside from the Veter project team, no one else seems to be advancing the state of robot mechanics.

We’ve seen robotic chassis and hardware from the Veter team before, and this new version brings a whole lot more to the table. While the camera. GPS, compass, and ultrasonic sensors are the same from the previous build, there’s a whole lot more software inspired by [Sebastian Thrun]‘s autonomous car class to make this build a little more capable.

While the Veter team is using a Beagleboard for their on-board computer, it should be possible to change the hardware over to a more economical Raspberry Pi. Even then, it won’t be a cheap build, but we doubt you’ll find a better robotics platform for less.

Comments

  1. dALE says:

    I would like to take this time to welcome out future robotic overlords.

  2. HC says:

    It’s probably due to cost, but I’m surprised the rocker-bogie suspension system isn’t used in more designs.

    • dALE says:

      Rocker-bogie suspension offers better ground clearance by using the stub axle, are less prone to tipping because there is no spring load, can operate on a steep angle, and in some situations can take on much harder obstacles than other suspension systems, like how the Mars Rover’s six wheel design allows one set of wheels to be pressed against an object and lifted by the other sets of wheels. It also offers the ability to alter the center of gravity which is great for stability.

      More widespread use should lead to lower cost.

      • Cyberteque says:

        its not rocker bogy, its a Rover 5 chassis and it looks like they got sick of the tracks the way I dis and went to wheels

        • dALE says:

          Oh, my mistake. It still sports most of the same features, save the ability to climb over taller and steeper obstacles.

          • cyberteque says:

            once I went ti LiPo batteries and 2 dual channel motor controllers my Rover 5 was able to climb quite steep, loose slopes.
            with the 6 AA batery holder fitted out with 2800mAh batteries it wouldn’t even move at anything under 40-50% PWM.
            After I went to LiPo’s it now can crawl along at 5-10% PWM.

  3. Senso says:

    Looks like the Rover 5 chassis:
    https://www.sparkfun.com/products/10336

  4. WA says:

    So throwing a 3D printed body over someone else’s commercial chassis is considered ‘progress’ in robot hardware? Or is it simply, the creation of a physical robotic platform as the ‘progress’…

    Since both concepts have been done repeatedly by others, I am unclear as to what progress this project represents…

    • dALE says:

      Buzzword Alert! It is a platform that allows you to focus on the meat and potatoes of your project and not waste time reinventing the wheel.

      here’s a quote from their site:
      “Cloud-robotics and distributed autonomous robotic systems are promising future directions. Our platform is the step towards this direction and our customers could benefit by reusing our software and hardware and concentrate on their areas of competence.”

      When compared to Lego Mindstorms they have this to say:
      “Our system offers more computational power and more flexible set of software building blocks to solve typical robotics problems.”

      Their target audience is:
      “researches in robotics, artificial intelligence and computer vision as well as hobby robotics enthusiasts.”

      It appears as though the “Progress” comes in the form of the key requirements outlined on their site:

      *should be complete open (hardware and software)
      *equipped with typical and widely used set of sensors
      *easy customizable to integrate new or different types of sensors and actuators
      *energy efficient yet powerful on-board computer
      *provide bi-directional communication link to transmit sensor and control data in real-time
      *set of software modules which support distributed data processing and provide hardware abstraction layer. It should let developers concentrate on experiments and applications of their core competences
      *considerably lower cost comparing to the similar available products

      -dALE

      • dALE says:

        Note that Veter never used the word progress to describe what they were doing, that was put in the summary by H@D

        • WA says:

          Yes, and that was the point of my comment. All too many of these posts by H@D are written as if the authors have no knowledge of the subject they are writing about–which sadly seems to be the case. There have been numerous examples of commercially available robotic platforms for research purposes for years (nearly as long as robotics research has been occurring)… There is nothing new here, which is not to say that the product shouldn’t have interest, just that the article itself is poorly written.

          Unfortunately, those who right these posts (which like this one sound like press releases) seem to lack even basic knowledge of the fields they are writing about. Much like one expects from MSM, not blog posts by enthusiasts (which is what H@D is supposed to be)… This phenomena is bad enough when they are writing about something reasonably obscure, but this type of robotics is very mainstream and they should have known better.

      • WA says:

        Explain how any of that applies to the original articles claim, “we’re constantly surprised at the lack of progress in robot chassis.” There is absolutely nothing new about the chassis–after all it is a repackaging of an existing commercially available chassis. The rest of your comment also refers to electronics/software, also not something that makes sense when the original article starts out with

        No matter how many advances in electronics we find, we’re constantly surprised at the lack of progress in robot chassis. Sure, it should be a simple task to make a capable robot that looks cool, but aside from the Veter project team, no one else seems to be advancing the state of robot mechanics.

        • dALE says:

          You know, it doesn’t apply to it at all, the article claims chassis and hardware development as well as being the only group making progress, but then their website claims that they are simplifying hardware and using more complex hardware while taking advantage of the open source community for development…

          so… I guess they are just bringing in all these elements to one package with some 3dp components…

          Buzzwords like Progress are pretty misleading when used incorrectly.

  5. airuno2l says:

    “aside from the Veter project team, no one else seems to be advancing the state of robot mechanics”

    Looks like they’ve just strapped some hardware to a DAGU rover http://www.pololu.com/catalog/product/1551
    Not really advancing mechanics if you ask me.

  6. Saulius says:

    Take a look at this one. It’s open source and has android application with video streaming: http://lukse.lt/uzrasai/2013-01-carambola-powered-robot-v2/

  7. It reminds me of the sidekick “Joey” from “Beneath a Steel Sky”. I wonder if it can come up with some witty banter also?

  8. Kevin Keith says:

    It seems like it would be fairly easy to vacuum form something like this with hardware store polycarbonate. No 3D printer required (and probably better quality).

  9. Gert says:

    I’ll grant that it is nice looking. Beyond that, not particularly intriguing – aka all the hyperbole is annoying. Well done, nonetheless.

    • Andrey says:

      It seams like the 3D designer has made such a good job that most of the discussion are immediately moved in the body/mechanic domain. But, when you are saying “not particularly intriguing” did you also took in account what is “inside” of this pretty looking body?
      In particular, there are:
      * 4 sonars, light sensor, two video cameras, GPS receiver and pan-tilt compensated compass.
      * BeagleBoard xM which opens much wider range of possibilities compared to the typical micro-controller-based design or even slow ARM based boards.
      * Tuned Linux distribution (based on Angsgrom) with all required drivers, support for native compilation, etc.
      * Linux driver for motor control where Xenomai real-time kernel extension is used to improve latency and jitter. The driver implements precise PWM generation with GPIO and low-latency wheel encoders interrupt handling. In addition there is a PID controller for wheel rotation speed implemented using integer arithmetic.
      * Remote (over the Internet) access to all on-board sensors and actuators to support the currently popular Cloud Robotics paradigm. Remote access could be implemented using almost any modern language and could be run on on-board computer (for completely autonomous mode or on the remote computer (or in Cloud) to support calculation-intensive decision-making and sensor data processing algorithms.
      * Adaptive h264 video streaming with very low latency and real-time adaptation for the variable channel bandwidth. It is absolutely necessary for manual control mode over the Internet (otherwise it is only possible to drive very slow and it is extremely inconvenient).
      * “Driver” console application using modern 3D visualization techniques (OpenGL+GLSL shaders, VBOs, etc.). This application shows real-time video and sensor data from on-board sensors (including maps downloading from openstreetmap.org based on received GPS coordinates)
      * Series of localization and control examples from advanced “Artificial Intelligence for Robotics” class.
      * Example on using OpenCV for real-time image processing for autonomous navigation.

      And many other goodies. Everything, including CAD models for 3D printing, electronics design and software is open source and is freely available.

      Do you think it is not particularly intriguing?

  10. Joejoedancer says:

    Are people going to be upset when biotechnology makes all this robot stuff irrelevant?

  11. david bigsbe says:

    Using awesome too much makes you sound stupid.

  12. strider_mt2k says:

    This is all well and good, but I’m wondering if this can be applied in the field of evil robotics?

    Evil robots just don’t seem to be getting enough development.

  13. morganism says:

    you guys are aiming way to low here, try this

    http://qntm.org/robots

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 92,307 other followers