Grant: Alliance Grants

In a groundbreaking project, researchers are leveraging deep reinforcement learning (DRL), a type of artificial intelligence (AI), to turn complex mathematical challenges into fun and interactive games-a process dubbed as "gamification". Inspired by the success of DRL in outperforming humans in Atari games for example, this project ambitiously ventures into the realms of transportation and energy sectors, striving to revamp the operations of ride-hailing services and streamline home charging systems. The initial phase of the project focuses on an emerging field spurred by the surge in popularity of ride-hail services like Uber and Lyft. With the anticipated influx of autonomous electric vehicles in these platforms, there emerges a necessity for a centralized system capable of adeptly managing these fleets within defined geographical regions. The initiative intends to utilize real data from New York City's ride-hail and taxi ventures and the National Renewable Energy Lab, to train and test DRL algorithms on a game version of this problem. Conversely, the booming interest in electric vehicles has highlighted a significant challenge concerning the rise in home charging activities, which are expected to exert a considerable strain on power grids. This issue, when transformed into a game format, exhibits a Tetris-like structure, providing a novel approach to address the problem through gamification with DRL algorithms. As the research advances, it aims to scrutinize the versatility of gamification approaches to new problem characteristics and mitigate the various challenges for gamification to become mainstream. In the grand scheme of things, the project aims to devise solutions that can substantially aid in shaping a more eco-friendly and energy-efficient society.