Obsessive. Compulsive. Inquisitive. I know a little about a lot and a lot about a little.

Jul 12, 2016

Stealing Creativity

It’s a painfully obvious imitation, boarding on counterfeit. I’d actually call it a counterfeit, but it’s not as good as the original. My latest project is something I’m calling “The Grabnar”.

The idea came about after I saw a snippet of a video in which a team of college students were showing off their latest project. It is a 3D printed tool to collect regolith on the Moon or Mars. It looks like they had squished the bottom half of a ball, then cut that ball into five equal sections. The sections open up as the tool is lowered, and draw regolith into them as it is raised. Apparently they based their design on the sea urchin. So they stole from nature, I stole from them. They want to further science. I thought it looked cool.

The biggest area my design standouts from the original is the method of opening and closing the doors. They used a threaded rod, or a lead screw. I’ve gone a servo to accomplish the same tasks. Granted, this limits my range of motion, doesn’t give me as much finesse and likely isn’t as strong. But it looks cool, which is what this project was all about.

Iteration 1. It worked. Kind of. More or less. Not that well.

Iteration 2. Two big changes here. The first was swapping where the inner and outer plunger connected to the scoop. The second was the arm. It was made to be much more symmetric than the first arm. This design worked great.

This is the third iteration. There is a threaded tube you can turn to operate the claw instead of just pushing the inner plunger up and down. The threads on the outer plunger run counter to the threads on the inner plunger. Turning the thread piece (upper left corner) opens and closes the doors. The arm also got a redesign to make it easier to print.

This is the final assembly, motorized an all. It’s basically the same as the picture at the top of the post. But with, you know, a servo.

Unlike my previous project, this worked on the first iteration, more or less. I didn’t copy their design exactly, but if you look at the two side by side, there’s no denying where my inspiration came from. Maybe that’s the reason for it working from the start. It’s also on the order of magnitudes easier to design than, say, a clock.

My secondary goal, behind just making something neat, was to explore fused filament fabrication (also know to most as 3D printing). I wanted to be able to print all of these parts on their own, without any additional support material. I found that presents an entirely different dimension to designing an object. But that constraint, like any constraint, can trigger creativity, and is reflected in the end product. There are a few overhangs that don’t turn out perfect when printed, but on the whole everything printed well.

This project could actually use nothing but 3D printed parts, with 3mm filament rivets to hold it together.

My printer uses 1.75mm filament and I had 3mm metal rods around, so I went in that direction. I used 3mm E clips to make sure the rods don’t slip.

The Grabnar is more or less complete. I’m still thinking through how to make piece displayable. Maybe turn it upside down. The doors sort of look like petals at that point. I could add a motion detector and turn the Grabnar into a mechanical, 3D printed Venus Flytrap. We’ll see.

Biomimicry

This whole experience has gotten me fascinated in mechanical biomimicry. Or, I should say, stoked my already existing fascination with biomimicry.

That curiosity was started by (the now seemingly ubiquitous) Strandbeest. I could do a whole post on Theo Jansen’s creations. Not from an engineering, or even intelligent point of view. Just in a that-excites-me-like-I-am-a-kid-getting-my-first-Nerf-gun kind of way. The way the PVC constructions move so seemingly natural and fluidly is awesome in the truest sense of the word.

The Grabnar, inspired by a project inspired by nature, in and of itself doesn’t exactly have that flow, but I’ve been working on making it seem to move with some kind of intelligence, or at least instinct, through the code that controls it servo.

And that exposes an interesting dichotomy of mechanical biomimicry. You can have an object whose locomotive design itself is what produces the appearance of life. Or you can have an object that in and of itself doesn’t move “naturally”, but given the proper instructions via code, can appear lifelike.

The Grabnar relies on both. As does another project I’m working on. Details on that in a future post.

What Jansen does with those two thoughts is fascinating. He combines them. Using a learning algorithm, he devised the ratios for the legs on his creations that would give them the most lifelike appearance. Their code is their DNA. Though they need only the wind for locomotion, they were born of a computer program.

If you want to hear ideas of biomimicry, Strandbeest and code as DNA, you should really check out Jansen’s talk at the MIT Media Lab.