designing a double-sided flip cut puzzle

We've been working on the design of the first puzzle for the axon tier! Originally, we were thinking of doing a puzzle with our wood inlay experiments but we've hit some roadblocks with that and are postponing the inlay work for a later puzzle. Basically we kept trying to make the inlay finer and finer to get higher image fidelity but that causes a high failure rate and maybe starts to undermine what we were trying to do. Time to switch gears! We're making a double-sided puzzle using our wave cut algorithm inspired by a cardboard puzzle we learned about at the Puzzle Parley conference this summer.

Its a double sided puzzle cut from both sides so you can’t use the laser kerf (the thickness of the line cut by the laser) to figure out which side is up. A trick for solving most double-sided puzzles is to see where the cut is thick versus thin to distinguish the sides (the thick side was face up in the machine when it was cutting). If you look closely at the second image you can see that some of the lines are thick and some of them are thin. That’s because we cut half of the lines from one side and then we flipped it over and cut the other half of the lines from the other side. This was inspired by these impossible cardboard puzzles that came out in the 1980s where they would do a similar thing with a diecut cardboard manufacturing method. See the puzzle pictured above "World's Most Difficult Jigsaw Puzzle" manufactured by Buffalo Games.

Jules made a prototype of the idea first using his coral cut algorithm on a painting by Van Gogh. The idea was just to make the puzzle super hard due to not being able to easily distinguish the sides of the puzzle but what I really liked about it was how the cut lines read more as long linear waves instead of a grid. The laser cut thickness became a design element we can play with. We decided to play into that by using our wave cut and the famous Hokusai image "The Great Wave" for the puzzle.

To make cutting the puzzle from two sides possible, we generate our puzzle off a rectangular grid. Usually we don't use rectangular grids but this let's us split the cut lines easily into 2 sets of lines: "verticals" and "horizontals." To make the puzzle we cut the horizontals then flip it over and cut the verticals. The overall effect on the waves on the first prototype was good but the pieces themselves felt repetitive. So the next step was distorting the grid along one direction into waves...and then inevitable distorting it the other direction.

We're still working on this cut. But I wanted to share these in process sketching so you can see what goes into designing a puzzle. Apologies in advance for any typos. I didn't have a chance to proof read this.