• aditya mehrotra.

chip updates: some issues and the plan moving forward [updates]

So today was a real pain for a few reasons. None of the control structures we have tried have worked thus-far. Probably because we've been approaching them in the wrong way.

While I agree, the best model will allow us to achieve the most optimal solutions sometimes. This isn't the case all the time. Engineering isn't about creating the best model, it's about picking the model that works for the situation. Now here's the thing. I think what we need to do instead is come up with a much simpler way to control the leg, not one that involves so much math. The amount of math we are doing at the moment is making the leg slow, and it's barely even working. Not to mention when we tried actuating it the rear left leg ripped its cables right out.

Frankly spending so much time on math is making this project not fun - we just need a simple control structure. One that works. And I hope to come up with that tomorrow.

The next few days we'll make a really really simple model of the leg and try to get it to follow some sort of walking path/trajectory.

Friday when the parts come to fix the rear left leg, we will use them, for now we will disable the rear left leg.

We're going to come up with the simplest possible model that works for all four legs that allows us to actually move the leg in whatever configuration we want to move it in the space we define. We want to create and equation with a closed-form solution and it may mean going back to the 2D simplification and just trying to control it differently.

Idk, we will see tomorrow. We really have two weeks to do this.

Remember, engineering isn't about how complicated can we make the problem. It's about how simple can we make it.

#simplicity #controls

0 views0 comments
© copyright 2019 | aditya mehrotra