Embedded Files
A collaborative effort of S. J. Tian*, S. Banerjie*, A. Dave, P. Typaldos, and A. A. Malikopoulos
A collaborative effort of S. J. Tian*, S. Banerjie*, A. Dave, P. Typaldos, and A. A. Malikopoulos
Project Overview
Project Overview
This project is a collaborative effort of the Information and Decision Science Laboratory (IDS) at Cornell University to develop a virtual reality (VR) simulator for autonomous driving with large language models (LLMs) to incorporate passenger preferences. This project re-imagines the control framework of Connected and Automated Vehicles (CAVs) to a passenger-centric formulation, where language input provided by passengers is used to align the CAV's driving style in aspects of comfort, speed, aggressiveness, etc. Subsequently, this project verifies the effectiveness of the proposed framework through VR demonstrations.
Motivation for LLMs in CAV Controls
Motivation for LLMs in CAV Controls
LLMs are large pre-trained transformer models [1] that are capable of consistently carrying out complex language processing and code generation tasks across a variety of applications. As a consequence, they can enhance CAV controls by interpreting natural language commands in the context of contextual data, such as road conditions or weather reports. This property can be used to improve the overall decision-making of a CAV in real-time, leading to transportation that adapts to the needs of passengers without compromising safety and efficiency.
Motivation for VR in CAV validation
Motivation for VR in CAV validation
Effectiveness of a control approach can be verified through a combination of simulations and real-world trials. In VR simulations, various scenarios can be tested in a controlled environment, measuring response accuracy, decision-making speed, and overall safety. These simulations also allow for the collection of subjective passenger experience ratings, providing valuable insights into user comfort and satisfaction. VR experiments have the salient advantages when validating CAV control approaches by ensuring low costs, quick turnaround times, and guaranteed safety of participants.
Meet Our Principle Investigator
Meet Our Principle Investigator
Andreas A. Malikopoulos, Principle Investigator
Andreas A. Malikopoulos, Principle Investigator
Dr. Andreas Malikopoulos is a Professor in Civil & Environmental Engineering and Director of the Information and Decision Science Lab at Cornell University. He was previousely a Professor in Mechanical Engineering and Director of the Sociotechnical Systems Center at the University of Delaware. He has also held roles at Oak Ridge National Laboratory and General Motors.
His current research focuses on optimizing cyber-physical systems to enhance energy efficiency and reduce emissions. He has received numerous awards throughout his career, including the IEEE Intelligent Transportation Systems Young Researcher Award and the Alvin M. Weinberg Fellowship. He also serves as an Associate Editor for Automatica and IEEE Transactions.
Contact: amaliko@cornell.edu | Room: 324 Hollister Hall | Phone: (607) 255-4734
Meet Our Project Team Members
Meet Our Project Team Members
Simon Tian
Simon Tian
VR Product Manager, CARLA | Python | VR Developer
M.S. Systems Engineering
About Information and Decision Science Lab (IDS)
About Information and Decision Science Lab (IDS)
The overarching goal of the IDS Lab is to enhance understanding of large-scale, complex cyber-physical systems (CPS) and establish rigorous theories and algorithms for making CPS able to realize how to improve their performance over time while interacting with their environment. The emphasis is on applications related to emerging mobility systems (e.g., connected and automated vehicles, shared mobility), sociotechnical systems, social media, and smart cities.
In the IDS lab, we believe that diversity of age, experience, race, ethnicity, religion, and gender can contribute to creating a dynamic intellectual group. We are committed to creating a diverse and inclusive environment of mutual respect and one that rejects discrimination, prejudice, and intolerance.
*Equal Contribution
References:
[1] A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, and I. Polosukhin, “Attention is all you need,” arXiv.org, 02-Aug-2023. [Online]. Available: https://arxiv.org/abs/1706.03762. [Accessed: 02-Aug-2024]
Image Courtesy of IDS Lab
Author of the Page: Simon Tian
Edited By: Aditya Dave
Page updated
Report abuse