AUM2020 Global Workshop: Session 9: Modelling urban activities

Duration: 2 hours 10 mins

Share this media item:

Embed this media item:

<iframe width="_width_" height="_height_" src="https://sms.cam.ac.uk/media/3420104/embed" frameborder="0" scrolling="no" allowfullscreen></iframe>

Choose size:

About this item

Available Formats

About this item

AUM2020 Global Workshop: Session 9: Modelling urban activities's image

Description:	1. How do car-lite policies impact housing mobility choices? A LUTI microsimulation analysis Mr Rounaq Basu and Prof. Joseph Ferreira (Massachusetts Institute of Technology) 2. Spatio-temporal Demand Modelling for On-Demand Transit Services Mr Nael Alsaleh and Prof. Bilal Farooq (Laboratory of Innovations in Transportation (LiTrans), Ryerson University) 3. Sensitivity analysis of housing market simulation Dr Xiaohu Zhang and Prof. Joseph Ferreira (Massachusetts Institute of Technology) 4. Temporally explicit models of firm behaviour Mr He He and Prof. P. Christopher Zegras (Massachusetts institute of Technology) Convenor and Discussant: Prof. Joseph Ferreira (Massachusetts Institute of Technology) Host: Dr Ying Jin (University of Cambridge)

Description:

1. How do car-lite policies impact housing mobility choices? A LUTI microsimulation analysis
Mr Rounaq Basu and Prof. Joseph Ferreira (Massachusetts Institute of Technology)

2. Spatio-temporal Demand Modelling for On-Demand Transit Services
Mr Nael Alsaleh and Prof. Bilal Farooq (Laboratory of Innovations in Transportation (LiTrans), Ryerson University)

3. Sensitivity analysis of housing market simulation
Dr Xiaohu Zhang and Prof. Joseph Ferreira (Massachusetts Institute of Technology)

4. Temporally explicit models of firm behaviour
Mr He He and Prof. P. Christopher Zegras (Massachusetts institute of Technology)

Convenor and Discussant: Prof. Joseph Ferreira (Massachusetts Institute of Technology)
Host: Dr Ying Jin (University of Cambridge)


Created:	2021-02-16 12:57
Collection:	Martin Centre AUM2020: Modelling the New Urban World
Publisher:	University of Cambridge
Copyright:	The Martin Centre
Language:	eng (English)
Keywords:	AUM2020; Online Global Workshop; Martin Centre; Architecture; Modelling the New Urban World;


Abstract:	1. Transformative technologies like automated vehicles, and emerging services like mobility-on-demand and ridesharing, are changing the ecosystem of urban mobility. Land use-transport interaction (LUTI) models provide appropriate platforms to test the impacts of such services on cities. Although these services are purported to have mixed effects on cities, there is a general consensus that these services will increase accessibility. We approach the ‘car-lite’ policy through this lens of increased accessibility, and base this study in the city-state of Singapore. Different study areas are chosen in a manner similar to the differences-in-differences approach, in order to tease out the effects of initial neighborhood vacancy rate, vehicle-free behavior, and tight markets on policy impacts. We also design different scenarios that represent varying market reactions to the policy, and compare them to a baseline where the car-lite policy is never implemented. Study areas that are initially less ‘tight’ (i.e., have higher vacancy rates and lower vehicle-free rates) are found to have significantly larger transitions to vehicle-free behavior. Additionally, our finding of accessibility-induced gentrification speaks to the importance of considering the endogeneity in housing and mobility choices while formulating policies that may seemingly feel relevant only to the transportation realm. Providing appropriate mixes of housing typologies with adequate affordable housing, in addition to restricting car use for higher-income car-owning households, are suggested as strategies for designing car-lite neighborhoods. 2. In this study, we use the operational on-demand transit (ODT) data collected from Belleville's pilot project to perform temporal analysis of the users' waiting time, fleet size, and the trips distribution, develop origin and destination patterns, and investigate the relationship between the demographic characteristics and the ODT trip production and attraction levels. Moreover, we present trip production and distribution models for the ODT service using four machine learning algorithms. Based on our findings, we further provide some useful policy recommendations to the operators and municipalities for sustainable planning, design, and operation of new as well as ongoing ODT projects. 3. This work is thus conducted to examine the sensitivity of the housing market portion of SimMobility—an agent-based micro-simulation platform of land use and transportation interactions that simulates daily housing market bidding. Using a calibration for Singapore, it confirms that the model can be calibrated with reasonable parameters that lead to dynamic equilibrium under constant demand and supply conditions. The market response is sensitive to the ratio of supply and demand but relatively insensitive to initial conditions. 4. We present a new dynamic microsimulation model of firm hiring and firing decisions. Our disequilibrium approach explicitly models hiring and firing rates, allowing us to simulate employment expansion and contraction over time and space. We derive a tractable likelihood function for estimation of the model and conduct a parameter recovery exercise with synthetic data to verify the feasibility of the approach. We are now in the process of collecting datasets to apply the model to the Greater Boston Area.

Abstract:

1. Transformative technologies like automated vehicles, and emerging services like mobility-on-demand and ridesharing, are changing the ecosystem of urban mobility. Land use-transport interaction (LUTI) models provide
appropriate platforms to test the impacts of such services on cities. Although these services are purported to have
mixed effects on cities, there is a general consensus that these services will increase accessibility. We approach the
‘car-lite’ policy through this lens of increased accessibility, and base this study in the city-state of Singapore. Different
study areas are chosen in a manner similar to the differences-in-differences approach, in order to tease out the
effects of initial neighborhood vacancy rate, vehicle-free behavior, and tight markets on policy impacts. We also
design different scenarios that represent varying market reactions to the policy, and compare them to a baseline
where the car-lite policy is never implemented. Study areas that are initially less ‘tight’ (i.e., have higher vacancy
rates and lower vehicle-free rates) are found to have significantly larger transitions to vehicle-free behavior.
Additionally, our finding of accessibility-induced gentrification speaks to the importance of considering the
endogeneity in housing and mobility choices while formulating policies that may seemingly feel relevant only to the
transportation realm. Providing appropriate mixes of housing typologies with adequate affordable housing, in addition
to restricting car use for higher-income car-owning households, are suggested as strategies for designing car-lite
neighborhoods.

2. In this study, we use the operational on-demand transit (ODT) data collected from Belleville's pilot project to perform
temporal analysis of the users' waiting time, fleet size, and the trips distribution, develop origin and destination
patterns, and investigate the relationship between the demographic characteristics and the ODT trip production and
attraction levels. Moreover, we present trip production and distribution models for the ODT service using four
machine learning algorithms. Based on our findings, we further provide some useful policy recommendations to the
operators and municipalities for sustainable planning, design, and operation of new as well as ongoing ODT projects.

3. This work is thus conducted to examine the sensitivity of the housing market portion of SimMobility—an agent-based
micro-simulation platform of land use and transportation interactions that simulates daily housing market bidding.
Using a calibration for Singapore, it confirms that the model can be calibrated with reasonable parameters that lead
to dynamic equilibrium under constant demand and supply conditions. The market response is sensitive to the ratio
of supply and demand but relatively insensitive to initial conditions.

4. We present a new dynamic microsimulation model of firm hiring and firing decisions. Our disequilibrium approach
explicitly models hiring and firing rates, allowing us to simulate employment expansion and contraction over time and
space. We derive a tractable likelihood function for estimation of the model and conduct a parameter recovery
exercise with synthetic data to verify the feasibility of the approach. We are now in the process of collecting datasets
to apply the model to the Greater Boston Area.

Available Formats

Format	Quality	Bitrate	Size
MPEG-4 Video	640x360	674.53 kbits/sec	642.26 MB	View
WebM	640x360	439.68 kbits/sec	418.65 MB	View
iPod Video	480x360	483.58 kbits/sec	460.44 MB	View
MP3	44100 Hz	250.37 kbits/sec	238.40 MB	Listen
Auto *	(Allows browser to choose a format it supports)