GoPiGo Dreams, Promises, and Reality

I admit it - I am a GoPiGo3 fan. I feel like it delivers everything it promises.

  • It allowed incremental investment (low entry fee, supported expansion).
  • It expands on the huge Raspberry Pi functionality and support system.
  • It provides versatile sensor interfaces (analog, digital, serial, I2C, USB).
  • It provides multiple programming language support.
  • Even multiple development modalities are supported.
  • The developers are accessible.
  • It supports a community forum.
  • It provides open source examples and drivers through github.
  • The platform is very well engineered for durability and expansion.
  • The platform is very well documented
  • I like the small size.
  • I like the user selectable rechargeable power concept that allows me to keep limits on what Li-Ion I allow in my home. (Only iPads, iWatches, iPhones, iRemote, and Nikon camera batteries allowed.)
  • The power system is robust and supports 3rd party sensors and effectors.
  • I have over two years invested in learning about the GoPiGo3 and I still love it, and have not out-grown it.

Ok, why am I saying this?

ClicBot on KickStarter.

For $200, $300, $400, $500, $600 the imagination runs wild.

  • It would appear to come with a few synthetic emotions, which I have been very intrigued by since first reading about “Braitenberg Vehicles: Experiments in Synthetic Psychology, 1986”
  • It appears to have built in vision sensor support
  • It appears to be beautiful, wonderful, functional
  • It appears to allow a paradigm shift beyond basic robot functions to personality and human interaction.
    .
    Promises, Promises …
1 Like

@cyclicalobsessive,

Yes, the GoPiGo3 is a great learning tool for learning to use Python to control a robot chassis.

However, it has its limitations.

As an example, can it eventually be programmed to do this simple task?

The physical limitations of the GoPiGo are obvious when it comes to being used for something other than a learning tool/programming experience.

Regards,
TCIII

2 Likes

Wow, thanks. That is a great write-up.

I have been watching the PiWars “Eco Challenge” bots, and now this LitterBot DonkeyCar, and yes, of course, I’m not sending my GoPiGo3 to the beach any time soon.

BUT generating a training set for how to drive around the house, avoid walls and furniture, and return to his charging doc, processing the training set on my desktop, and then running the result on my GoPiGo3 bot will be both a learning tool/ programming experience and a “useful” function for my Carl.

This only confirms that my GoPiGo3’s limitation is me (and a 3D printer).

1 Like

@cyclicalobsessive,

I have built a modified Nvidia Jebot using a Lynxmotion 4WD chassis and a PCA9685 Servo Controller/L298N Motor Driver combination in place of the recommended Adafruit DC MOTOR + STEPPER FEATHERWING ADD-ON so that I could drive DC motors bigger than what is used on the Jetbot or the GoPiGo3 chassis.
Take a look at the examples on the Jetbot GitHub to get an idea of what Jetbot can do in comparison to GoPiGo3.
Regards,
TCIII

2 Likes

Every component certainly has its advantages and limitations for a specific requirements set.

At the moment the GoPiGo3 still is the best platform for Carl and me.

1 Like

Totally agree. Everything it promises, and then some. The promises are only limited by your imagination.

Very possibly, given sufficient coding skill - which I don’t have at this point.

Oh, and BTW, that’s not such a “simple” task. It’s a lot like walking. It LOOKS  easy, but I can guarantee you that it’s not the trivial task it seems.

Absolutely true in every respect.

Every time I have failed to reach a goal I have set, it’s because I  failed to follow through, or psyched myself out of doing it.

Sure, some tools make certain tasks easier - and yes, a Jetson Nano beats the living [poop] out of a Pi, even a Pi-4, when it comes to complex AI tasks - but does that mean the Pi can’t do them? absolutely NOT!  It just takes longer.

Even though a shunting locomotive is a better tool for the job, you CAN pull a freight car with a VW Beetle. (Of course, a fully loaded, hundred-car coal train might be a bit more than it can handle. . . :wink:)

@thomascoyle11859 has done an excellent job of pointing out the shortcomings of the GoPiGo, and yes, a Jetson Nano runs rings around it - then again a Jetson Xavier runs rings around the Nano. . . And a fully tricked-out PC with a dozen or so processor cores, a half-million or so CUDA rendering pipelines and liquid nitrogen cooling beats the living bejeasus out of all of these.

But then again, where can you spend a couple of bills on a full-up 'bot and have a total HOOT of a time? Jetson Nano? At least for a total neo like me, that’s a bit chewy.

I for one, don’t expect my GoPiGo to be able to run an autonomous vehicle on the Los Angeles Freeway, if for no other reason than my coding skills have not yet reached that pinnacle of excellence.

However, something I can  do, and do it in Bloxter, is code Charlie so that he can wander around the living room without smashing himself face-first into something. I’m still working on that project, but have accomplished enough of the individual pieces that I absolutely know that it’s achievable, even by someone with my limited skills.

I, like @cyclicalobsessive, honestly believe that the only thing that limits what you can do is YOU.

This is what I use as my e-mail signature - I think it says it best.

What say ye?

1 Like

Oh, and by the way. . . this is right off the kickstarter page:
(Note the highlighted prices.)

I don’t think this   'bot is going to be driving around picking up trash anytime soon.

You can get a full-up GoPiGo for less than two bills. Of course, if you want to trick it out, that’s your problem!  (Yea, I should have such problems, 'eh?)

1 Like

@jimrh,

Since I am a System Engineer I leverage off of what other experts have produce in the way of programming, circuitry, and mechanisms to create something unique and useful in the Real World.

If I wanted to become proficient in Python for robotic programming then the GoPiGo3 would be a great way to learn. However I want to create something that can be used for specific purposes in the Real World and I feel that GoPiGo3 is very limited in that aspect. The chassis is not designed to carry much of a load and its battery power is limited compared to other robotic chassis.
My version of the Jetbot is a 4WD chassis with a 5000 ma-hr battery for the four motors and a 13 amp-hr battery to power the Jetson Nano. It can easily patrol the outside of my house all night without having to return for a recharge.
Real World projects should be the end goal and GoPiGo3 is just not the vehicle to get me there.
Regards,
TCIII

2 Likes

Ok, I’ll bite.  What do you mean by a “real world” application?

Of course, I totally respect your desire to “Go Boldly Where No Nuclear Physicist Has Gone Before”, and if the GoPiGo won’t get you there, then by all means, get what you need.

I bought myself a Nano to play with - with the emphasis on “play” with.  Maybe I can interface it to the GoPiGo chassis and make it do useful things.  Maybe I cam play around with AI, or graphics programming on it - later when I acquire the necessary skills.

I have to admit that the main reason I bought the Nano when I did, was with the idea of plugging it into the GoPiGo - and that’s turning into a bigger nut to crack than I had anticipated.

I guess what got my goat was, (what seemed to me), the total dismissal of the GoPiGo as an unworthy “toy”.  Maybe it IS a “toy” - at least in your eyes - but for me it still represents a significant challenge.

Unlike both yourself and @cyclicalobsessive, there’s a lot for me to learn.  I’m cutting my teeth learning from existing software - both the GoPiGo control panel app, (maybe I can get it to work with TKinter?), and the Remote Camera Robot code that I hope to be able to interface an actual joystick to.  I’m learning a lot about Git, Source Control in general, and a lot of the pesky “real world” things that go into making a 'bot of any complexity do what I want it to do.

But that’s me.  I would love to be able to sit at your feet and learn what you have the privilege to know, as best as I can.  In the meantime, I guess I’ll go play with Charlie and learn what he can teach me.

1 Like

While many hobbyists do enjoy the GoPiGo3, it is first and foremost an educational robot, designed for school use (middle school and high school).

#differentGoals

2 Likes