The Self-aware City Massachusetts Institute of Technology, June 2006 © 2006 Andres Sevtsuk. All Rights Reserved. Abstract This thesis explores the idea of real-time urban space management. While increasing amounts of real-time information about the city, specifically the location of people and resources, appear, it becomes necessary to explore how different strategies of distributing real-time location information can be used as urban design tools for a more sustainable resource allocation. I focus on the study of street-parking, a system that clearly has a market situation with demand and supply, but due to lack of information is poorly managed today. I argue that an equilibrium state of the parking market, similar to many other urban space markets, is a constant over demand. The important challenges are therefore optimizing allocation and managing queuing. I propose five different strategies of using real-time location information to reduce search times and analyze the system through computer simulations and logic. Borrowing ideas from Game Theory, I try to illustrate how collaborative behavior between drivers could yield most efficient results from both the individual and the group point of view. Lastly, I outline some challenges that the use of real-time information systems introduce to the realm of urban design in general. Thesis supervisor: William J. Mitchell |
||
Parking Simulations: (source codes to all simulations are under the title link) 1. Traditional Model...................................................... 2. Intel_1 Mode.......................................................... 3. Intel_3 Model .................................... 4. Intel_5 Model..................................... 5. Intel_7 Model......
Optimal Meeting Point Simulations:
1. Sum of Individual Meetings ........................... 2. Group Meeting....................................................
|
||
![]() |
||
![]() |
||