Show HN: A Modern Palletization App | Hacker News Reader

Show HN: A Modern Palletization App

github.com

153 pointsvld542 comments

When searching on the internet for these type of apps, I didn't find many that were open-source AND easy to use. A lot of them had complicated interfaces, although they had loads of features as well.

So what I had in mind when making Stack Solver was an app with a modern interface that has the most essential features. It is also well integrated with Microsoft Excel and renders a 3D customizable drawing.

Stack Solver is programmed in C# using the WPF framework to ensure it is fast and light. The interface is built using WPF UI, a library that allows it to keep up with modern trends (specifically the Fluent design).

It is a work in progress with tons of new features planned and it is my first "serious" project so I would appreciate any feedback :)

Open articleView on HN

Ah! That's a pretty much orphan subject yet the world is constantly moving pallets.

For pallet stability you would like to have a convex perimeter so the stretch wrapping maintains the boxes effectively. And also criss-crossing boxes for shear resistance.

But you need to align as much boxes corners as possible to get vertical stiffness.

A solver proposing stacking patterns with these constraints would be outstanding.

Does center of mass matter? Or not so much.

Should lighter boxes be higher, so they don't get crushed? Or again, doesn't matter.

Yes right, I'm biased toward single product pallet. Then high mass items should be placed on the first lower pallet, low mass on the pallet put on top of the first one in the truck. The solver starts to be complex.

> But you need to align as much boxes corners as possible to get vertical stiffness.

The magic might not be obvious?

I've made an animation[0] using the demo image[1] from VladM7's Stack-Solver.

[0] - https://img.go-here.nl/palet-stack.gif

[1] - https://github.com/VladM7/Stack-Solver/blob/master/img/scree...

pallet is 120 long. 21+21+21+21+35=119 pallet is 80 width. 21+21+35=77

> And also criss-crossing boxes for shear resistance.

Mirror the next layer.

Besides useful it also looks impressive to have a full looking pallet with criss-crossing.

And if it can spit out an OpenRadioss input file, one could launch pallets (almost) directly.

Looks great! If you did do something related to truck fill, would be great to see if it could be generalised to 3D bin packing.

I’m currently working a free / open source warehouse management system so great to see other people working on supply chain problems!

I would maybe like to contribute to WMS. Wrote a WMS for a company I work for.

I use .net, C#, SQL, blazor.

Would love to hear more about a FOSS WMS - which aspects are you focusing on initially?

Amazing! I’m trying to make a WMS that’s very easy to customise (rather than configure) - will come with a highly opinionated set of base processes but then will be built in a way that it’s incredibly easy to edit them (rather than building a million parameters/toggles).

Handset processes are installed like apps.

Would you be interested in learning more about the ITAD industry? I built a warehouse monitoring solution that included quiet a bit. Surveillance, sales-performance tracking, employee-performance tracking, guided repair on broken machines removed from pallets, cost-benefit-analysis "what do I work on" kinda apps. We supported 50+ orgs in my local community and some major fortunre 500s before i left over some ethical concerns. Schools, hospitals, and other refurbishers/resellers/recyclers in our industry some years back all used our software.

This sounds interesting! Have you published it yet?

Not yet but will shoot you a message when I do!

I would appreciate that! Please do!

Stuff like this is great.

Fast moving consumer goods is a vastly underrated represented technology sector, they do so much cool stuff that just goes completely unnoticed.

I started my first job in a warehouse and it really gave me a good look at the proprietary technology powering day to day life. It should be more open but few work on stuff like this.

I work in this field, as a developer for a proprietary warehouse management system.

The entire industry runs off closed source systems with little to no accessibility for external developers. The companies developing these systems like it this way, as it means they get a large number of billable hours through their service departments, since there are very few developers available with the expertise to work on these systems.

I've had some friends in the logistics industry ask about a mobile app for this. How hard would it be to convert that to mobile? Are you depending on some complex libraries for the solver, or did you implement the algorithm yourself? I have 0 to no knowledge of C#.

I implemented the algorithm myself and although I don't have much experience with mobile apps programming I think that converting it to other programming languages or adapting it for other frameworks would be fairly easy. The real problem is the 3D rendering which is done using WPF 3D (Windows only), so I would need to completely rework it.

If you're going to rework it, perhaps consider making it a web app? Wrap it in Electron for desktop use, I bet there's something similar for iOS / Android.

That'd make it available as both a dedicated app for power users, and an (ad-supported) variant for occasional use.

Moving to Avalonia from WPF should be relatively straightforward (except, possibly, the WPF 3D part), and Avalonia is capable of targeting mobile platforms (even though it really is desktop-first, when compared to Uno or MAUI). As a bonus it will run on macOS and Linux too.

Even if mobile v1 was simple level by level text output, it'd probably still be useful.

Probably yes. Implementing only the text output should be very simple. Anyways, I'll look into it.

I suppose this would be useful for large supply chains. Once you get away from that and move to shipping from warehouse to store, though, this would be a largely be a waste of time. Stacking a pallet with various different sized boxes and shapes is a bit of an art form that is largely learned through trial and error until you "get" it (I was on a full time night stock crew while I was in college). It would literally take more time for workers to refer to the layout than it would for them to simply stack the goods onto the pallet.

Pallet stacking can also be done automatically (with robots), and there are solutions (only privately developed/owned to my knowledge) that calculate optimal stacking for heterogeneous loads - interlock boxes for stability, weight (distribution & total), height, etc. Each box is characterized on induct (weighed, scanned, sometimes manually updated with metadata on crushability/fragility). Then an algo figures out best way to put the boxes on a pallet. I've seen this for grocery, but applicable in other use cases.

It's only time-consuming for workers to follow the planned layout when those workers are human.

Just lasted 2 days moving pallets barrels and cylinders. I needed the customers to accept non-full 300 lb barrels instead of the full 500 lb. Couldn't do it. Pallets and wrapping were to load a bit easier but often we cut the shrink wrap at the first stop and the pallet had stuff for like 3 stops. Seems like the shrink wrap was practical to keep stuff from flying around but also regulatory, cops could write a ticket.

And yet people are allowed to load and drive their own rented U-Haul vans.

Thanks for this! We ditched our Tops Pallet software a year ago and have been looking for something similar. Where can we donate to the development?

I just set up a donation page, you can find it on the github repo page, on the right. But please keep in mind that for now I have to finish my exams so you may have to wait a bit for new features. Also, any feedback and suggestions for improvement are appreciated especially since you work in the field.

We considered using Tops Pallet software at one point. What made you ditch it?

Amazing job, love seeing native apps like this rather than webapps. Keep it up!

Robot palletizers use such algorithms, but do humans need a program for this?

Another case of "Machines should think. People should work."

Absolutely. Pallet arrangements are usually a required spec from the customer to the vendor. The arrangement order matters considerably for shipping and needs to be agreed upon beforehand.

I tried to make some sort of joke or pun based off of shipping containerized software, I couldn't make it work. Oh well.

Only thing I think it needs is the option to change the UOM from metric to imperial.

Not something I have a use for, but that's pretty cool.

here's my alternative 3d bin packer https://3dpack.ing

The humble pallet is the red blood cell of civilization.

There's a ton of really cool optimization ideas like this in Industrial Engineering - factory production schedule optimizations, path optimizations, lots of knapsack problems. What a treat that there are people willing to pay for this kind of work.

Nice analogy!

I agree, very nice.

I'm a truck driver. My analogy, zooming out a bit, is that the entire road system, at least the National Network, is a system of interconnecting and overlapping conveyor belts.

The trailers are trays on the belt.

https://en.m.wikipedia.org/wiki/National_Network

Shipping containers are another genius-level invention - even better than wheels on luggage.

They used individual crates before!

Fun fact: inside many of the trailers you see, it's just manually loaded and stacked, and manually unloaded, individual boxes. No pallets.

Not most. But many.

Another fun fact: US/NATO logistics is pallet-based. Russian is largely not.

Unloading shells on a hot day in a "bucket brigade" must be fun.

True - still a lot better than painstakingly offloading a ship's cargo onto trains and trucks and vice versa, one crate at a time. All of civilization runs a good deal faster and better because of shipping containers

Reply on news.ycombinator.com