MegaBots wants to have a real-life mech battle. Preferably without dying. So far it’s not looking good.
When I first heard about the MegaBots Kickstarter project last year, I definitely giggled like a ten year old. Anybody who wants to build big robots and make them fight is okay by me.
We wrote about it here. I sponsored the project. I proudly hung the poster on my office wall, where it serves as a daily reminder that robots, when fighting, are, in fact, awesome.
Well friends, it just got real. MegaBots has come a long way. The production value alone is an impressive improvement over their earlier videos, and while the test dummy most definitely (spoiler alert) DIES A HORRIBLE DEATH, the bot, overall, faired reasonably well given the level of abuse dished out. Have a look:
First question. Couldn’t they have anticipated the paintball breaking through the diamond grate? Aside from just seeming obvious, it seems like there would be much quicker and easier ways to (a) anticipate the problem, (b) test possible solutions, and (c) iterate on the results. Let’s see, there’s:
1) Math, fellas. Good ol’ F equals m of a. You’d know immediately that your little welds won’t hold against that blow.
2) Testing. Why test on the whole bot when all you need is a single window? Make a single rectangle of your mesh, weld it to a frame, and shoot it in the lab. If you want theatrics, put your test dummy behind the window. Much cheaper, much faster, and…
3) Iterating on the results. If you can build ten different window designs and try them all without needing the full bot, you’ll find better solutions more quickly.
I was also disappointed that they only took the wrecking ball to the dead front of the bot. Shouldn’t they also hit it from the sides and back? And what about shooting the side of the cockpit?
I’m also interested to know where the weak points are in the design. What are the critical joints, and how are you protecting them from impact, torsion, and/or friggin’ lasers?
Engineers! I summon thee! How could these guys test this sucker more efficiently?