This project developed an simulation of a an intersection routing fully autonomous vehicles through it using wireless communication. This research was completed in conjunction with Ford to help quantify the impact the cyber attacks might incur on autonomous traffic flow.

This project developed an modification of typical adversarial attacks against Deep Reinforcement Learning (DRL) agents. It trained several DRL models in different MuJoCo environments and then attacked them using our proposed method.

Although mostly static, this website was constructed in Nuxt to continue to grow my familiarity with it. It uses the new view-transition api to provide smooth animations between pages. It uses UnoCSS to construct rules for dynamic classes for the view transitions.