OpenNest 2: Struggling to Find Optimized Results

Grasshopper OpenNest
1

Hi everyone,

First, I’d like to extend my thanks to Petras for creating OpenNest and for the incredible contribution this plugin makes to digital fabrication. It’s an amazing tool, and I really appreciate the effort that’s gone into its development.

That said, I’ve been struggling to find an optimal arrangement for my nesting problem and was hoping the community might have some advice. Here’s an example of the result I’m getting for this case:

image
image1111×513 19.3 KB

I’ve experimented with various parameters—adjusting the seed, the number of iterations, the population size, and the mutation rate—but the results don’t seem to improve significantly. In particular, I’m curious why this last piece is left out of Sheet 2, even when I’ve allowed for 4 rotations and run 800 iterations.

Interestingly, I can achieve a straightforward solution across three panels manually, like this:

image
image1108×286 19 KB

However, I haven’t been able to tweak the parameters to produce an optimal solution on just two panels:

image
image1130×331 22 KB

At this point, I’ve been working around the issue by manually refining the arrangement, starting with the best result I can achieve after a few iterations. I suspect there’s something I might be overlooking in how I’m setting up the plugin.

I’ve attached my Grasshopper definition and 3DM file for reference. If anyone has insights or suggestions on how I could improve the results, I’d be very grateful!

Thanks again to Petras for such an incredible tool, and thanks in advance to anyone who can help.

Thanks again for all the help,

OpenNest2 Optimizable.gh (25.6 KB)
OpenNest Optimized.3dm (1.3 MB)

2

I guess the thing that these 4 parts need to be arranged like this in order for all the parts to be succesfully nested into 2 sheets is a very big drawback

image
image1488×369 18.6 KB

all the random solutions the solves generates where 1 is not fitted around 2 like that, and where 3 is not fitted around 4 like that will just create heat :slight_smile: and if you consider the probability of that happening on its own by just dragging the seed slider then… my guess is that we are gonna wait a lot before that happens :slight_smile:

this said, if instead of those 4 parts I nest their 2 bounding rectangles , at seed number 4 I get this:

image
image1763×641 106 KB

nesting_Re.gh (25.7 KB)

for the general case, whenever I have parts to be nested which need to be combined in a particular way in order for the nesting to be successful, I try to nest them already-combined in their final-configuration, and let the nesting engine take care of the placement of their already-combined shape

3

Thanks @inno
Your proposed solution makes sense. Thanks for sharing and taking the time.
I was hoping for an automated solution with the objective to fully integrate the workflow without having manual intervention: Configurable Model to Toolpath generated (Bark Beettle).
Is the 3 pannel solution then the best one can hope for on an automated solution in a ‘reasonable time’ ?

4

Hi,

I have fixed a few bugs on OpenNest version 2.4.0 please give a try.

In previous version genetic algorithm was not working properly.

Tight arrangements are tricky, but at least it will improve over number of iterations.