Actions
Forum Thread Details
  • Replies 16 replies
  • Subscribers 2564 subscribers
  • Views 2384 views
  • Users 0 members are here
Related

Why Are You Waiting to Apply for the Beaglebone AI RoadTest

rscasny

A famous person once said, "The Future Is Now."

image

So, why wait to the future, what you can do now. Like, apply for the BeagleBone AI Roadtest: BeagleBone AI

 

Built on open source Linux, BeagleBoneRegistered AI fills the gap between small single board computers (SBCs) and more powerful industrial computers.

 

BeagleBoneRegistered AI makes it easy to explore how artificial intelligence (AI) and machine learning can be used in everyday life. Machine learning is supported by an optimized TI Deep Learning (TIDL) software framework and pre-installed software tools. Here's a few specs:

 

  • Works with existing BeagleBoard.orgRegistered BeagleBoneRegistered cape add-on boards and many available third-party cape add-on boards and enclosures
  • 1GB RAM and 16GB on-board eMMC flash with a high-speed interface
  • USB Type-C for power and superspeed dual-role controller; and USB Type-A host
  • Gigabit Ethernet, 2.4/5GHz WiFi, and Bluetooth 4.2/BLE
  • microHDMI video and audio output
  • Zero-download out-of-box software experience
  • TI SitaraTm AM5729 processor with: Dual 1.5GHz ArmRegistered CortexRegistered-A15 microprocessor subsystem 2 C66x Floating-Point VLIW DSPs
  • 4 Embedded Vision Engines (EVEs) supported by the TIDL software framework library for machine learning
  • 2x Dual-Core Programmable Real-Time Unit (PRU) subsystems (4 PRUs total) for ultra-low latency control and software generated peripherals
  • 2x Dual ArmRegistered CortexRegistered-M4 co-processors for real-time control and subsystem management
  • IVA-HD subsystem with support for 4K at 15fps H.264 encode/decode and other codecs at 1080p60
  • VivanteRegistered GC320 2D graphics accelerator
  • Dual-Core PowerVRRegistered SGX544Tm 3D GPU

image

So here's my question:

 

what's keeping you from applying to get a free Beaglebone AI to experiment with and get into the world of Machine Learning?

 

At this point it's a complete mystery to me? Kindly inform me, Watson!

 

 

 

 

 

 

  • The BBAI has incredible potential.  In fact, I can see it being my preferred card over all others for embedded, robust designs hands down....but...

     

    What's holding me back is it unfortunately doesn't have GPIO support out-of-the-box and I'm not smart enough to derive the device tree on my own.  In turn, my review would be very limited and not reflect its potential.

     

    If they release an update where I can read and write a couple of pins and generate PWM, I'm all in.

     

    -Sean

  • I have taken a good look at the Beaglebone AI and am very interested.  I also have an AI project underway that I am blogging about weekly which would be a good match for the Beaglebone AI but I am currently committed to doing it with a Raspberry Pi due to timing and familiarity with the Pi.  It might be possible to do a Beaglebone AI RoadTest following that project as a way to compare to the Pi which would be interesting.  However, the timing isn't great as the holidays will be upon us at that time.  I plan to have a think about it closer to the November 15th deadline and commit if I think I can dedicate the time needed for a good RoadTest.

  • A competitive RoadTest application is not a swiftly written one! It takes time to research the product, it takes time to learn about how to use it, its strengths/weaknesses and think of a thorough worthwhile (and perhaps unique) RoadTest.

     

    That being said, sometimes it takes time to do the write-up of the application - other times I might be sitting at a road-block for imagination. But if you have a better imagination than I do, I'd encourage all to apply!

     

    - Gough

  • I'm not waiting, I am in park mode. It isn't so much a lack of inspiration, it is just that all the AI applications I want to do require a huge amount of time. When I road test a module it usually includes a real application I am interested in. As with many of the really powerful modules, my implementation would take longer than a normal road test period and it would usually have a lot of risk of not working well.

  • in reply to Gough Lui

    Well said . Let me add my two cents worth. This is the best board for someone wanting to LEARN more about AI. You have the best documentation and one of the few plug-it-in-and-run setups around. Cloud 9 is simple and effective to do any task whether or not you know Linux. Need I say more? Anyone passing this up would be missing a great opportunity.

    Clem

  • My trepidation in preparing an application is due to a lack of core samples. Geologist use core samples to determine what riches lie below the surface. Can the return justify the investment ti mine the deposit? As & commentary indicates the foundation for a good application is defining an outcome.

     

    I have no Beaglebone experience. No core samples to draw from. I have founded a few applications on that supposition. My goal is to learn. I want to discover. The subject matter interests me. Unfortunately, these type of applications lead to introductory type RoadTest Reviews. I took it out of the box. I loaded the operating system. I got it to work. I'm done.

     

    I pre-screen myself from many RoadTests because of my simplistic approach. Maybe RoadTests can have a scale of deliverables? suggestion that the Beaglebone AI is a great AI introduction device really piqued my interest because of my desire to expand my knowledge curve.

     

    I take some pride in being selected as a RoadTest reviewer. I want the review to reflect that pride. That takes time and effort.

    Sean

  • I have some ideas in mind, but need more time to write it all down. Furthermore I'm working on another roadtest application as well. This makes it a bit tricky as you can not be selected for more than one roadtest at a time and there might be some preference for a specific roadtest. can you shed some light on this dilemma?

  • in reply to clem57

    Really cool stuff. And really need time to explore.

    This is good for Industrial Application Designer, Best of all BeagleBone Family.

  • The task is beyond my current capability.

    Plus I am generally underwhelmed by AI.

     

    DAB

  • Aren't you jumping the gun on this one... still 15 days to go (in some cultures that is almost a different generation away). I trust you're not planning to conduct "exit polls" and more "surveys" to gauge take up as we all know how "accurate" those can be...  or have your been bitten by some sort of bug - I got to read through +100 applications with RPI so let's see if we can top it with this one. imageimage

     

    It should be pretty obvious to many that this looks to be a fairly powerful product (and ahead of the curve like many of the other beaglebones). It just takes a bit of upfront effort reading the bumph and getting your head round how to tackle such a road test and whether you have what it takes.