How to efficiently override global state variables in a compiled Android APK for resource optimization?

I am working on a performance optimization project for a heavy 3D simulation app (similar to Truckers of Europe 3). My goal is to bypass the linear progression system to test high-end assets (trucks and skins) without going through the standard $O(n)$ grinding process.

The Problem:

The application stores its economy state and level progress in a local binary file. I want to implement a "Saturated State" where:

current_balance is set to INT_MAX.

player_level is forced to its maximum threshold.

Currently, I'm using a memory injection approach at runtime, but I'm encountering a Memory Management issue: when I unlock all 50+ high-res skins simultaneously, the heap usage spikes, leading to an OutOfMemoryError (OOM) on mid-range devices.

What I've tried:

I’ve successfully patched the current_balance variable in the data structure.

I attempted to use Lazy Loading for the unlocked assets, but the game's original engine tries to pre-cache all "available" skins in the UI list.

Code Snippet (Conceptual Logic):

Java

// Current approach in the optimization patch public void applyStateSaturation() { if (OptimizationModule.isActive()) { GlobalState.setBalance(999999999); GlobalState.setLevel(MAX_LEVEL); // Triggering the unlock for all assets AssetManager.unlockAllVehicles(); } }

Questions:

How can I intercept the engine's pre-caching routine to prevent it from loading all textures into memory at once?

Are there better patterns for State Overriding in compiled environments to ensure stability (similar to the patches used at HNHAYVL for v0.7.64)?

I’ve seen this implemented smoothly in some technical mirrors like https://hnhayvl.net/truckers-of-europe-3/, but I'm struggling with the memory overhead part.

Any advice on memory-efficient state manipulation would be appreciated!

Tags: android java optimization memory-management reverse-engineering