GAMMS Simulator

Overview

A game engine built in Pygame using Python for the USC GAMMS robotic laboratory. I am a CSCI Graduate Research Assistant assisting researchers from the University of California, San Diego, the University of Southern California, Department of Defense, and the University of Michigan in developing a simulator for robotics. I work with a team of 4 developers in building and supporting a graphics API.

Github: GAMMSim/gamms

This page will be updated as time passes.

Graphical Optimization

I introduced a dynamic and static layer drawing system into pygame. This works by converting the function calls into pixel arrays and streaming it as a texture. The result is, compared to older builds below, a almost 100x improvement in runtime!

Graphics and Visual Simulation - Early Development Build

Using NetworkX and Pygame, I helped develop an AI graphical interface to display real-world GPS data in an interactable environment on a PC screen. The UI is also real-time based on what the AI can see at that moment. This is an early development build.

Dynamic AI Visuals

The display is dynamic to what the AI can see. Because the AI has a sensor range, it means it can not see the whole map. In addition, this uses a custom-built render caching that allows the program to selectively display points without significant loss of frame rate.

Custom API

Another feature I helped develop is the ability to for developers to interface and customize the engine.