General Misc > Robot Videos

Rubiks cube solving robot

<< < (2/3) > >>


I left the whole robot part for others to solve; my software'll give you a list of the right values for each square, but making a bot that can read, write, and smoke is a whole different matter.

One of my metrics for quantifying where we are in robotics is what I call the "Bender Test." When we can build an autonomous humanoid robot, fueled by alcohol, which can survive indefinately by committing petty thefts, pawning the loot,  and begging to keep itself in "fuel,"  we'll be able to make a robot do anything. We've got a ways to go, but I have faith that we'll get there.  ;D

you could use rollers with numbers and encoders on them... :P

Well, if you're just making it a robotic sudoku board capable of solving itself, I'd say it's better just to make it pure software.  The board would be cooler, perhaps, but only marginally so, and not enough to justify the extra effort.  It sounded like this was a school project, but is it a programming project or a robotics project?

hardware based looks cooler, and offers other unique challenges outside of just software . . .

if you want simplicity, you could just solve it paper and pencil  :P

Granted, but there are far more interesting robots to be built I think; then again, I wasn't terribly impressed with the cube solving bot either; in both cases, the mechanical challenges are certainly educational, but they don't strike me as particularly impressive. Now, if they built the solver inside-out, by making a rubik's cube into a bot capable of solving itself, that would impress me! As would a sodoku solver that actually sat solving sudoku puzzles from a book, filling in the answers with pencil. Naturally, building the bot would be an educational challenge, but no more or less than any number of other robot projects. It's just not that interesting to me, but that's entirely subjective, and others no doubt feel the same about my roachbot project.

I actually do find the idea of writing a pure software sudoku solver interesting, and might play around with it some time; I'm sure there would be some delightfully subtle complexities I completely missed in my high-level algorithm above. But that would be a weekend project with no cost, where building the bot would take many weeks and $$$. Plus, people could use a software sudoku solver (if I try it, I'll probably make it web-based), while the bot would just be a novelty.


[0] Message Index

[#] Next page

[*] Previous page

Go to full version