Alex Wong

EECS Department, University of California, Berkeley

Technical Report No. UCB/EECS-2023-166

May 12, 2023

http://www2.eecs.berkeley.edu/Pubs/TechRpts/2023/EECS-2023-166.pdf

This paper takes a holistic approach toward managing autonomous vehicle traffic in parking lots. Parking in lots is a crucial aspect of navigation, and a potentially time-consuming one, with drivers often having to circle around multiple times or manage complicated intersections. Parking large fleets of vehicles can be especially slow, thanks to the amount of congestion created and lack of vehicle coordination. To improve this, we present an autonomous framework for improving parking lot efficiency for a wide range of scenarios, including fleet parking. First, we develop a path planning and collision avoidance framework for individual vehicles as well as a simulation framework for managing a large group of vehicles in a decentralized manner. Then, we turn toward fleet management and improving parking efficiency for incoming fleets of autonomous vehicles. In particular, we focus on the approach of intelligently assigning spots for incoming vehicles to minimize the fleet's time to park. We pair this work with an existing parking lot dataset and run extensive simulations testing spot assignment strategies, including random assignment, assignment using a neural network, and assignment using an pre-trained driver intent prediction model. Overall, we find significant parking time savings compared to human driving, and we identify scenarios where different spot assignment strategies could be utilized.

Advisors: Francesco Borrelli

---

BibTeX citation:

@mastersthesis{Wong:EECS-2023-166,
    Author= {Wong, Alex},
    Editor= {Arcak, Murat and Borrelli, Francesco},
    Title= {Improving Parking Lot Efficiency through Autonomous Control and Assignment Strategies: A Microscopic Traffic Simulation Analysis},
    School= {EECS Department, University of California, Berkeley},
    Year= {2023},
    Month= {May},
    Url= {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2023/EECS-2023-166.html},
    Number= {UCB/EECS-2023-166},
    Abstract= {This paper takes a holistic approach toward managing autonomous vehicle traffic in parking lots. Parking in lots is a crucial aspect of navigation, and a potentially time-consuming one, with drivers often having to circle around multiple times or manage complicated intersections. Parking large fleets of vehicles can be especially slow, thanks to the amount of congestion created and lack of vehicle coordination. To improve this, we present an autonomous framework for improving parking lot efficiency for a wide range of scenarios, including fleet parking. First, we develop a path planning and collision avoidance framework for individual vehicles as well as a simulation framework for managing a large group of vehicles in a decentralized manner. Then, we turn toward fleet management and improving parking efficiency for incoming fleets of autonomous vehicles. In particular, we focus on the approach of intelligently assigning spots for incoming vehicles to minimize the fleet's time to park. We pair this work with an existing parking lot dataset and run extensive simulations testing spot assignment strategies, including random assignment, assignment using a neural network, and assignment using an pre-trained driver intent prediction model. Overall, we find significant parking time savings compared to human driving, and we identify scenarios where different spot assignment strategies could be utilized.},
}

EndNote citation:

%0 Thesis
%A Wong, Alex 
%E Arcak, Murat 
%E Borrelli, Francesco 
%T Improving Parking Lot Efficiency through Autonomous Control and Assignment Strategies: A Microscopic Traffic Simulation Analysis
%I EECS Department, University of California, Berkeley
%D 2023
%8 May 12
%@ UCB/EECS-2023-166
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2023/EECS-2023-166.html
%F Wong:EECS-2023-166