That is so cool! (and “Masters Level” indeed. I dabbled with Braitenberg Vehicles on the GoPiGo3 which are “two neuron” simulations, but C. Elegans? Wow!
I did English proof reading for a German team using C. Elegans for flavonoid influence on cancer cells.
I couldn’t find the Python3 version of the Connectome tutorial. Have a link handy?
It’s now a Jupyter notebook and not a pure Python file.
Supposedly the pure python for GoPiGo3 is here:
I’m going to add this to my TODO: list - looks really interesting, and I think Dave always wanted “worms in his brain”…maybe that was some other of questionable intelligence.
Interesting to see John Cole in the author list!
if I remember correctly the pure python file wasn’t working with Python3. It was timing out or something. It was taking a really long time to get started. Jupyter Notebook handled it much better
My two centavos:
IMHO, learning robotics is difficult enough as it is, without adding layers of complexity or requiring massive external resources as a dependency.
If this were natively available, and were as simple as a Bloxter script - “move from here to there” - without requiring a massive “flight simulator” capable robot as a co-processing host, then it would be useful.
As it is, it takes a relatively complex process and somewhat steep learning curve and makes it both more complex and steeper than necessary.
Again, IMHO, this is like ROS - a layer that the GoPiGo doesn’t really need unless you want to explore more advanced robotic concepts than the current environment can support.
This reminds me of Bradly’s Bromide - “Inside every large, complex program is a smaller, simpler program struggling to get out.” If we can navigate to the moon with nothing more than a pocket calculator’s worth of computing power, surely we can move from one place to another without requiring the robot to carry a Cray on its back.
Note:
This does NOT invalidate current research. I am simply expressing my opinion as to the general applicability of this as a language model for the GoPiGo3.
What say ye?
P.S.
@cyclicalobsessive
@greenvh
I think both avenue’s of research have value; your emphasis on larger logic models applicable to more complex reasoning and my emphasis on logic models that are more simple and easier to understand for the less advanced student. It’s not that your model is wrong, I just think that I can achieve results that are substantially similar with a smaller, less complex, logical model.
The problem wasn’t that “it wasn’t working with Python3”, it was that it was expecting Python3 v7, and certain dependencies were incompatible with later Python versions.