I Redesigned StarRupture's Drone Rail Junctions
By The French Monkey · 2026-01-13 · SHOWCASE
I've been playing StarRupture for a couple of days now, and I want to be clear up front: it's a great game. Maybe too great, my factory keeps growing and it is beautiful. But somewhere between expansion projects, I ran into a problem with one of the game's core features: the drone rail junctions, the main way you transport items between facilities.
And once I saw the issues, I couldn't unsee them. So I did what any designer with a free weekend would do, I tore the models apart, figured out what went wrong, and redesigned the whole system from scratch. Here's the full breakdown.
The Gameplay Problems
Clipping instead of model switching
The height adjustment option on junctions is genuinely great. But when you lower a junction, the model just clips through the platform instead of switching to a model that fits that height. The game already lets you switch model alignment on most objects with a right click, why not do the same here? It would look far better, and if the game ever implements a proper floor building system, this clipping becomes an issue that has to be fixed anyway.
The clipping also causes a real gameplay problem: you can't place a junction on top of an already-built platform below it, because the model can't clip through existing geometry. So you have to delete the platform first, place the junction, then rebuild. That's a rough mechanic for an item you place hundreds of times.
One fix solves both problems: switch models based on the selected height instead of lowering one model through the floor.
Distribution networks eat too much space
When you're building factory lines with multiple inputs, there's currently exactly one way to build a proper distribution network, and it takes four blocks to distribute four different items efficiently. That's a lot of real estate for something so fundamental.
What I'd love to see is a junction stacking option. A multi-layered rail system would let players build dedicated systems tailored to their needs instead of being locked into one layout. More creativity, less sprawl.
The Design Problems
Beyond the mechanics, the junction models themselves are inconsistent in ways that bug me as a designer:
• The four-way cross-section base is different from the other three. Same category, different design. It looks nice, but why did this one get extra care while the others didn't? • The junction support looks out of place. The rail size is simply too small for the base it sits on. • The multi-rail got clamp supports; nothing else did. The clamps look cool. So why are the other rails just... sitting on top of the base?
Reverse-Engineering the Original Models
To understand the logic behind these designs, I started by rebuilding the base model myself. That's when things got interesting.
The top part of the base is really off, the line work doesn't make sense, and I basically had to cover it up by smoothing everything. As a designer, that tells me something: this beam asset was probably part of a different, larger model that got cut off at the top and kitbashed into the junction set. The design isn't bad. It just reads as unfinished, or in this case, amputated. That's not my opinion; that's what the lines say.
Going through each piece:
• Four-way crossing base: the beams are duplicated and rotated 90 degrees. Not a problem on its own, but it creates a lot of intersections in the model. It's lazy work. • Single junction: the size is wrong, and it just sits on top of the base, colliding with it. This clearly wasn't the first model made in this category. My guess: the smaller rail size caused technical issues in the engine, and they bulked it up to compensate. The rail itself is extruded and covered up, it's not built like the others, which smells like a last-minute change. • Three-way multi-rail: the only one with clamping supports, and the most finished design of the bunch. I genuinely like this one. • Three-way splitter: sits on the base just like the four-way. Given its size, they probably skipped the clamps because it's not noticeable. • The beam itself: I really like it. The forms, the cuts, the shapes, it feels right. But it fails at one job: it doesn't communicate height. You can't tell which elevation step you're on or where the lower step is. It needs a visible feature to guide the eye.
The Redesign
With all the issues catalogued, I knew exactly what the redesign needed. I rebuilt everything in Fusion, a couple of hours, several iterations and redraws later, here's what I ended up with:
A new base and beam with visible indentation. You can now read every elevation step directly off the beam.
Dedicated models for each elevation step, including floor level, which keeps a proper platform support so the beam isn't just extruding out of the platform. No more clipping in and out, no more deleting platforms to place a junction.
A universal clamp system that fits the beam design and accommodates all rail types. Simple approach: the clamps extend to fit the splitter and the multi-rail. No more rails awkwardly floating on top of bases.
A modular mounting system for multi-layered junctions. This is the part I'm most excited about. With just four mount models, rotatable, you can create over 260 variants, opening up real multi-line network design. It could even be an unlockable tier for a corporation. When mounted, the new rail sits one elevation step up, so proportionally it fits the current game as-is: place your junction at the height you want, select a mounted rail, hover over the top of the junction, place. Done.
A screen mount that attaches to the base at any height. Use case: showing what items are passing through the line, and maybe line issues or missing items. There's currently no way to know what's flowing (or missing) through a rail, and honestly, this would look great too.
The result is a junction system that's more satisfying to look at, more satisfying to build with, and fixes every issue on the list.
A Note to Creepy Jar
I'm not here to bash game developers. I know what making a game takes, I've started projects myself and didn't have the time to finish them. These things take years. If this were a solo dev with little to no budget, every issue above would be completely understandable. No questions asked.
But this game likely had a multi-million dollar budget and something like 50 people working on it. It took me two hours to reverse-engineer the original models and eight hours to build a better alternative. My redesign isn't perfect, it has flaws, and it would still need inspection and iteration to be engine-ready. But that's exactly the point: this is fixable.
So please, Creepy Jar, fix your junctions. They're the most-placed item in the game and the models we stare at the most. It's a shame, because almost everything else is great: the music, the sound, the landscape, the atmosphere, the combat. I'm not even going to mention the bugs and glitches, those will get patched. These junction issues aren't early-access issues. They're a core system that wasn't fully thought through from the start, and it needs a redesign.
