38C3

A Competitive Time-Trial AI for Need for Speed: Most Wanted Using Deep Reinforcement Learning
2024-12-28 , Saal ZIGZAG
Language: English

All challenges and achievements in creating a competitive time-trial AI in NFS:MW.


15 years ago, at the height of my eSports career, I uploaded an (unofficial) ESL record at Need for Speed: Most Wanted (2005) (NFS:MW) to Youtube. In the meantime Deep Reinforcement Learning became popular and ever since I have dreamt of creating a competitive AI for my favorite racing game of all time: NFS:MW. Now finally the time was right: The hardware is fast enough, good software is available, and Sony's AI research has proven the task is actually doable. Hence I thought: "How hard can it possibly be?".

This talk will present in detail all challenges and achievements in creating a competitive time-trial AI in NFS:MW from scratch - including but not limited to - hacking of the game to create a custom API, building a custom (real-time) OpenAI gym environment, steering the game using a virtual controller, and finally successfully training an AI using the Soft-Actor-Critic algorithm. All code including the API is written in Python and is open source.

Almost 20 years ago Sebastian was a professional player for Team Alternate aTTaX for Need for Speed. In the meantime he has completed his studies in economics and statistics and started a career in data science and machine learning and is currently heading a Data & AI team at E.ON.

Ever since reinforcement learning has come up, he wanted to build an AI to build his records! And he started that journey in 2023...