Graduate student Xin Wang will be defending his thesis.

• Date: Tuesday, June 1
• Time: 8 a.m.
• Where: Microsoft Teams, join here.

Here’s an abstract of the thesis:

Public and private transportation systems (Uber, Lyft, Juno, etc) represent a solution to the problem of human mobility. However, their massive adoption often results in traffic jams, congestion, and delays, the product of the massive utilization of few popular roads. When these transportation systems are used in conjunction with vehicular crowdsensing, a system that utilizes built-in vehicle sensors for collecting data, the problem of uneven distribution for data collection will occur, leading to high data redundancy and poor budget utilization. This research seeks to present incentive mechanisms for vehicular crowdsensing that encourage participants to deviate from their pre-planned trajectories in order to collect sensing samples located at any place of the city. The goal of the research is to reduce vehicular traffic in areas of more congestion, while at the same time enabling the use of vehicles and their on-board sensors as nodes of a mobile sensor network. This allows the vehicle owners to profit from data collection, at the same time allowing city planners to utilize the power of crowdsensing to monitor road infrastructure and other environmental variables as well as to optimize the use of city network. Through dynamically modifying the vehicles’ trajectories, sensing samples are able to be collected from regions otherwise unreachable by originally planned routes.

The problem is modeled as a non-cooperative game in which a set of vehicles equipped with sensors are the players and their trajectories are the strategies. Thus, the solution corresponds to a model in which expected individual utility drives the mobility decision of participants. Through efficient computer science and data science practices, we aim to implement the algorithms and evaluate its performance through extensive simulations.

For more information, contact Dr. Luis Jaimes.