Shop OBEX P1 Docs P2 Docs Learn Events
MS Robotics Studio 4 Beta and new Parallax Robot — Parallax Forums

MS Robotics Studio 4 Beta and new Parallax Robot

wjsteelewjsteele Posts: 697
edited 2011-10-08 19:12 in Robotics
Check out the new MS Robotics Studio 4 Beta. They even talk about a new robot called "Eddie" from Parallax!!! :-)

Announcement here: http://channel9.msdn.com/posts/A-Look-At-Robotics-Developer-Studio-4-Beta

Check out Eddie here: http://www.parallax.com/eddie

And here is the new RDS: http://www.microsoft.com/robotics/

The biggest thing that I see so far is the inclusion of the Kinect SDK even in the simulation environment... so you don't actually need a Kinect or even a robot to design and build your prototype robot!

Bill
«1

Comments

  • Cats92Cats92 Posts: 149
    edited 2011-09-18 07:21
    Your link to Parallax does not work
  • wjsteelewjsteele Posts: 697
    edited 2011-09-18 15:00
    Hmmm... It works for me ok.

    Bill
  • TtailspinTtailspin Posts: 1,326
    edited 2011-09-18 15:54
    Link works for me too..

    Mr Gates has all the coolest toys, Now he wants to play with propellers, Man that IS cool...

    -Tommy
  • BaneBane Posts: 41
    edited 2011-09-18 17:29
    This is absolutely fabulous!
  • ypapelisypapelis Posts: 99
    edited 2011-09-18 18:40
    It looks really cool, but am I the only one who sees a big disconnect between the complexity and overhead associated with MSRobotics studio and the simplicity and straightforwardness of the propeller? Also, once you hook up a whole PC on a robot, why would you need a propeller for anything any more?
  • GordonMcCombGordonMcComb Posts: 3,366
    edited 2011-09-18 20:34
    ypapelis wrote: »
    Also, once you hook up a whole PC on a robot, why would you need a propeller for anything any more?

    I don't know how Eddie is engineered regarding its implementation of the Propeller, but generally you don't want a non-real time OS controlling hardware like the Madeusa. Instead, you use the PC, with its much larger memory, RAM, and processing power, for things that a microcontroller are not good at.

    Complex vision analysis is a good example. I use a Webcam connected to a Windows laptop, and some low-level DirectShow coding to do machine vision -- basically turn each frame into a color bitmap, and analize the bitmap. (There are open source libraries as well; AForge and OpenCV among them.)

    Some friends of mine here in SoCal do something along similar lines, with an HC11-based microcontroller to do the low-level hardware interfacing, and Windows with C++ coding for the heavy lifting:

    http://leafproject.org/

    Robin has written some killer routines for Open CV enabling the robots to do some advanced mapping and navigation; Bruce (the guy standing beside the water heater on wheels) writes his AI routines in Lisp, and Alex does the C programming on the HC11.

    Note the way the laptop is strapped to a vertical column, with the keyboard and monitor on the same plane. This is actually a pretty cool arrangement. It might make for an interesting alternative for Eddie.

    BTW, Bruce, who is a neurologist by trade, enclosed his bot in a solid body because as I understand it his work interest includes using machine intelligence to interact with persons with certain disorders, among other things. I imagine the open frame "guts" of the robot might interfere with accepting it as a friend, or something like that. It looks a bit ungainly, but serves a purpose. The animation on the laptop screen is in real time, includes natural and very realistic speech synthesis, and is based on AI. None of these things you can do on a Prop. Or even an Arduino! :lol:

    -- Gordon
  • Tom CTom C Posts: 461
    edited 2011-09-19 04:35
    @ wjsteele
    I already have the Motor Mount and Wheel kit installed on one of the original Parallax wooden bases. I wonder if they plan to make the controller board available by itself for those of us who already have the hardware chassis and motors?

    @ ypapelis
    You miss the point of this exercise. The Propeller does the grunt work (motor control, sensor processing, etc) on the robot while the MS Robotics Studio performs the higher level tasks such as vision recognition, etc.

    Regards,
    TCIII
  • Mike4421Mike4421 Posts: 131
    edited 2011-09-19 20:37
    @ypapelis.... this is great news.... this is Big news!!!
    There are many things happening in the 32bit microcontroller world now from diffirent manufactures. What this does, it lets you program your propeller into a platform that offers you 3D simulation, a Visual Programming Language, and now support for Kinects, and XNA 4.0 (Microsofts Game Programing Platform in C# for Windows Phone, Zune, and XBOX360)!!!
    All this can be used to a Propellers programmers benefit.
    Yes I know... the communications are not simple like standard RS232 or TTL or USB; they have to be put on steroids with the CCR and DSS, but the benefits are great. I stayed on the BS2, then moved on to other 32bit microcontrollers; I hope you guys take advantage of this!! you can use the Eddie code as an example for your connections and expand. Its nice to see RDS and Parallax in a new project again. :)
  • Matt GillilandMatt Gilliland Posts: 1,406
    edited 2011-09-19 21:05
    Hey TomC-
    I already have the Motor Mount and Wheel kit installed on one of the original Parallax wooden bases. I wonder if they plan to make the controller board available by itself for those of us who already have the hardware chassis and motors?
    Yes!
    -Matt
  • Matt GillilandMatt Gilliland Posts: 1,406
    edited 2011-09-19 21:08
    this is Big news!
    Yes it is!
    and welcome to the Forums Mike4421 :-)
    -Matt
  • ypapelisypapelis Posts: 99
    edited 2011-09-20 03:49
    @tomc, @gordon

    I realize the issue of separating the functions of a full blown OS from that of a microcontroller. I also think it is wonderful news for parallax to offer such a project, especially given the markets it opens and the vast number of folks that will be exposed to the wonderful world of parallax products. However, I still have some concerns, that I am having a hard time expressing. Maybe they are due to my background maybe I am indeed off, so let me try once more (promise, that's the last time!)

    Over my 20+ year career, I have worked on numerous projects, large and small, that have had the need for low level circuits and/or microcontrollers as well as higher level functions assigned to an OS-based computer (often more, sometimes 10's of "computers"). I have also used PCs (big and small) and written probably millions of lines of code at that level. In almost all cases, there is a tendency to ignore the low level stuff, or better yet, treat them as replaceable black boxes with little regard for who builds them. Let me describe a few examples.

    I was on a team that designed several large simulators, some costing tens of millions of dollars. When needing to read sensors, we used added VME based boards in the chassis (running VxWorks), each board costing thousands and nobody blinked, including the people paying for the project; if anyone had proposed a propeller based I/O card, they would have been laughed off the team (mind you, I am not defending this, just conveying it). The point is not that a propeller based card can't do the job, it's just that when thinking at that level, you tend to avoid worrying about the low level staff and just 'buy' it.

    I have worked on projects that involve driving cars autonomously around city streets. The GPS/INS alone cost about $40k. A total of about 4 PCs provided the horsepower for navigation A/I etc. When the time came to design the various low level microcontroller boards to deal with hardware interfaces, controlling motors, rods etc., nobody cared who built them (including me), pololu, parallax, TI, National, whatever, just buy one (or 10 of them) and hook them up, now let's get to the improtant staff (on the PC running Windows or Linux). Part of this is due to the fact that the problem space is at a higher level, so it naturally minimizes the low level staff.

    The parallax world has always been about grassroots, efficiency, attention to details, and focus on circuits/firmware. Somehow the equation changes when you get a full blown computer involved because the low level microcontroller becomes less important and replaceable. The kinect can be read directly through a USB card, most cameras can be read directly through USB, so most of the algorithms that people will focus on will be running on a PC (and require floating point and tons of horsepower), so what does the prop do? like someone said, control the motors; but so can tons of other products, so it looses its uniqueness.

    So I guess, what I am trying to say is that "we" care that the prop has 8 cogs, runs deterministic code and can detect pulse widths down to 12.5 nS, but somehow all of this becomes less important when you launch visual studio.
  • Tom CTom C Posts: 461
    edited 2011-09-20 05:00
    Hey TomC-

    Yes!
    -Matt

    Matt,
    That's great news to hear. I think that I spend most of my time building and modifying Propeller Proto Boards to meet my robotic requirements. Being able to purchase a drop-in controller will greatly reduce the time to get a functioning robot working!

    Regards,
    TCIII
  • GordonMcCombGordonMcComb Posts: 3,366
    edited 2011-09-20 09:42
    I wouldn't worry about Windows or any other "big box" operating system subsuming embedded microcontrollers, even when they're paired up together. One complements the other, and it comes down to what provides the lowest cost. It's cheaper to add an MCU than to beef up a host PC, and safer than relying on the OS to handle the kind of real-time low-level processing a robot requires.

    It's also easier to program. The OS handles functions in the abstract, but it's the MCU that handles the details. Earlier in this thread I posted an open source project that's similar to Eddie, but fully homegrown, pairing low-end Windows laptops with a custom HC11-based microcontroller board. It's hardly the first time it's been done, and won't be the last, and yet microcontrollers still appear to be the favorite way to directly connect to hardware.

    -- Gordon
  • Jorge PJorge P Posts: 385
    edited 2011-09-20 11:42
    MSFTnew on twitter just tweeted

    Scientific American: We're getting close to a robot in every home, especially after reporter met #EDDIE. http://bit.ly/pbM58e

    http://twitter.com/#!/MSFTnews
  • jdoleckijdolecki Posts: 726
    edited 2011-09-20 12:19
    So there is a service already written to communicate with the Propeller processor?


    thanks john
  • Tom CTom C Posts: 461
    edited 2011-09-20 13:10
    Hey TomC-

    Yes!
    -Matt

    Matt,
    Any chance of posting a shot of the new Eddie Propeller Controller Board?
    Regards,
    TCIII
  • Tom CTom C Posts: 461
    edited 2011-09-20 16:00
    Matt,
    Are you using an inhouse developed IR range finder or a commerically available one, like the Sharp IR modules, on the Eddie Propeller Controller Board?
    Regards,
    TCIII
  • wjsteelewjsteele Posts: 697
    edited 2011-09-21 14:16
    jdolecki wrote: »
    So there is a service already written to communicate with the Propeller processor?

    John, yes, MSRS includes all the functionallity to work with the Propeller powered Eddie robot. (Including a simulated version of the robot itself!)

    Bill
  • Matt GillilandMatt Gilliland Posts: 1,406
    edited 2011-09-24 08:05
    Any chance of posting a shot of the new Eddie Propeller Controller Board?
    Regards,
    TCIII
    Of course :-) Here ya go...
    104_4945.jpg

    -Matt
    1024 x 768 - 103K
  • Matt GillilandMatt Gilliland Posts: 1,406
    edited 2011-09-24 08:16
    Are you using an inhouse developed IR range finder or a commerically
    available one, like the Sharp IR modules, on the Eddie Propeller Controller
    Board?
    Hi Tom-
    Yes, they're the Sharp modules. They're becoming a standard stocked item, available individually or as part of some of our kitted robots too.

    We've got an acrylic mounting bracket for it too - you can see it on Eddie, and I'll post some closeups of it shortly.

    -Matt
  • ercoerco Posts: 20,256
    edited 2011-09-24 09:10
    Yes, they're the Sharp modules. They're becoming a standard stocked item, available individually or as part of some of our kitted robots too.

    Hey, I love them Sharp sensors, but aren't they discontinued? Where are they listed in your store? I know you have an old app at http://www.parallax.com/dl/docs/prod/acc/SharpGP2D12Snrs.pdf, but since then, ain't seen nuttin' honey.
  • Matt GillilandMatt Gilliland Posts: 1,406
    edited 2011-09-24 09:18
    Erco- I'll quote myself:
    ...becoming a standard stocked item...
    note the word 'becoming'. Apparently we found a source - don't know where though - Purchasing Dept. hates me, so we don't talk much ;-)
    I'll post pics and more info on Monday!
  • Tom CTom C Posts: 461
    edited 2011-09-24 09:51
    Hi Matt,
    Thanks for the info on Eddie Propeller Controller Board and the IR sensors that you are using. Will you provide the Sharp IR module part number on Monday?

    It looks like the motor controller power transistors (FETs) are well heatsinked! Did you decide not to have to fan cool the power transistors (FETs) like on the HB-25?

    I love the HB-25 and have never blown one out!

    Any ballpark idea on the pricing of the Eddie Propeller Controller Board?

    Will IR sensor mount be available separately like the PING mount?

    @erco
    I bought a bunch of GP2D12s from Lynxmotion and eBay a while ago so I am well stocked up. Lynxmotion still has the GP2D12s for sale: http://www.lynxmotion.com/p-260-sharp-gp2d12-ir-sensor.aspx Also includes a prewired cable! Went digging through my goodie box and found that I have at least 12 of the GP2D12s plus one GP2D120!

    Regards,
    TCIII
  • ercoerco Posts: 20,256
    edited 2011-09-24 10:51
    @Tom: Yep, I have my secret source, too! They're a bit cheaper at http://www.junun.org/MarkIII/Store.jsp

    I think they are discontinued, though. Might get rarer & more expensive as time goes by. Stock up now while you can! My fave Hamamatsu P5587 is history, even at Junun. Both of these are far more sophisticated than simple phototransistor sensors hawked elsewhere in the forum.
  • Tom CTom C Posts: 461
    edited 2011-09-24 17:41
    erco wrote: »
    @Tom: Yep, I have my secret source, too! They're a bit cheaper at http://www.junun.org/MarkIII/Store.jsp

    I think they are discontinued, though. Might get rarer & more expensive as time goes by. Stock up now while you can! My fave Hamamatsu P5587 is history, even at Junun. Both of these are far more sophisticated than simple phototransistor sensors hawked elsewhere in the forum.

    Hey erco,
    Thanks for the tip. The cable is only $1.10 too! Lynxmotion includes it for free, but charges $5 more than MarkIII.
    Regards,
    TCIII
  • Mike4421Mike4421 Posts: 131
    edited 2011-09-25 00:39
    Erco....Thanks for the Sersor code for Basic Stamp2, I'm going to start learning the Propeller chip; cant pass up this great project. Parallax has a great community and service.
  • graffixgraffix Posts: 389
    edited 2011-09-25 13:22
    Is there a price on the new board?Or estimate?
  • Tom CTom C Posts: 461
    edited 2011-09-25 17:59
    Matt,
    My present Parallax motor mount platform is the old wooden style, but is still very serviceable. The present 12vdc 7ahr battery fits fine underneath the platform. How did you manage to squeeze a 12vdc 14ahr battery in the same space?
    Just a thought.
    Regards,
    TCIII
  • Matt GillilandMatt Gilliland Posts: 1,406
    edited 2011-09-26 21:32
    Is there a price on the new board?Or estimate?
    Not quite yet, we're still figuring all that out. Will let you know as soon as we know :-)
    :-)
    -Matt
  • Matt GillilandMatt Gilliland Posts: 1,406
    edited 2011-09-26 21:37
    How did you manage to squeeze a 12vdc 14ahr battery in the same space?

    We didn't squeeze in a 14ahr battery. We simply rotated two 7.5ahr batteries by 90 degrees, connected them in parallel, increased the battery tray width, and cut it out of 1/4" Delrin (!) - it's nearly indestructible...

    Testing videos to come - we gonna drop it, we're gonna crash it, and we're gonna ride it :-)

    -Matt
Sign In or Register to comment.