Surrender? Never!

A project log for The Emergence Project

An experiment in group machine intellegence (Or lack thereof...)

Greg DaneauGreg Daneau 09/09/2014 at 03:220 Comments

It's been a couple of weeks since our last update. Some may have thought the Emergence Project would fade away into the night after we didn't make the cut for the Hackaday Prize. HA! This was in the planning stages in one way or another for years, and we're gonna see it through. Let me assure all True Believers that our coverage gap was caused by the Labor Day holiday, and an unexpected burst of activity at the annoying (but still essential) day job.

We're still working with Johnny on the light seeking behavior, but it's a slow slog. We've basically been working with him using the Scientific Method. We change X in his programming, then go into the back stairwell that has no windows, turn the lights out, and wave a flashlight in front of his nose to see how he reacts. Then we go back into the code, restore X, and change Y. We're slowly learning what parts of the code are responsible for which behavioral aspects.

Frankie, on the other hand, is as disobedient as ever. We put together a semi-permanent circuit board for him, which worked fine the night we installed it. When we turned him on the next day, however, he refused to move, and when we picked him up to see what was wrong his voltage regulator was almost hot enough to light a cigarette off of. We instantly disconnected the power, but it's painfully obvious that we have a short somewhere in the maze of solder joints.

However, we still have faith in the design and don't plan to change it. Johnny and Frankie have had enough successful run time to demonstrate that they do work as envisioned, with the inconsistancies explained by use of plug in breadboards and a string of software bugs.

So, while we fuss with our original two, the Great Parts Gathering for the final swarm has begun. The processors are on the way, and we have our casters and motor driver boards, as well as 13 of the 15 solar cells. We found the cells at a Tiger Direct brick and mortar location, and cleaned them out. The last two are on the way.

We have made some subtle changes, though. We went with a different motor driver, which is identical in all but name, but from a different manufacturer at a few dollars less a copy. We've also changed our main processor from the Ardweeny to the Arduino Mini Pro. It's a shade cheaper, and has the benefit of a couple extra I/O locations. With the extra versatility, we can include a small LCD display to read the bots' settings at a glance, instead of having to plug in and suck it out of their brains. The extra cost is being mitigated by the dollars we're shaving off here and there with the other components. Our original budget for each 'bot was $50, and our present projection is putting the actual cost at around $56. Not too bad, considering that every other robot we've ever built has usually come out double what was projected!