Open access
Author
Date
2021Type
- Doctoral Thesis
ETH Bibliography
yes
Altmetrics
Abstract
Urban transport is a substantial issue in many countries. Congestion and public transport unreliability need to be addressed. Any new policy needs to be tested comprehensively against people's behavior. This thesis comprehensively captures travel behavior and willingness to pay for the transportation modes. This thesis further develops an agent-based model of Greater Jakarta to model all complex interactions in which individuals interact with households, activities, and facilities.
There were several steps to develop the model. Firstly, the demand was built based on the survey of JICA in 2012 and combined with the RP survey to capture the trip purposes. We used the Bayesian network and generalized raking for the population synthesis. This method can generate the travel demand that represents the population in Greater Jakarta. We created a 30 million synthetic population based on responses from around 600,000 individuals from 170,000 households. Secondly, networks and public transport lines were built as supply facilities. The data from Open Street Map was used to create the network, and the data of public transport lines were scraped from Trafi. This supply and demand-side were used for running the agent-based model using multi-agent transport simulation (MATSim).
The thesis also created a state of the art a stated preference (SP) and revealed a preference (RP) survey with 5,000 respondents. This thesis developed a model by pooling SP and RP datasets using the multinomial logit model (MNL) and mixed logit model (MXL) to see how those modes affect mode choice. The parameter estimated from the model is used to calibrate the MATSim model so that the mode share from the MATSim model matches the HTS. Utilizing parameter estimated from the choice model claims to speed up the simulation time and convergence.
There are new modes of transportation options available that have been investigated, such as on-demand transports (car-based, motorcycle-based) that exist in many countries due to the development of information communication technology (ICT). Other alternatives will be available in the coming years, such as autonomous vehicles and urban air mobility (UAM).
The results show interesting results. First, the development of the scenario for greater Jakarta required a lot of effort. The availability of data and the population's size required expensive computations starting from population synthesis to model simulation. Second, we investigated the willingness to pay (WTP), e.g., the value of travel time savings (VTTS), value of travel time assigned to travel (VTAT), and the elasticity for all mode choice alternatives, including on-demand transport (ODT) and UAM. Third, we found that the recorded travel behavior answered several questions, such as who the users are, when the users use it, the users' purposes, and the speed at different distances and locations. Finally, we developed a scenario for policy recommendations using an agent-based model. In this case, we modeled road pricing implementation on main roads in Jakarta.
This research contributes by building an agent-based model scenario and investigates the impact of congestion charging in Greater Jakarta. This research will be a base of future research for developing more complex scenarios. This thesis advises on the improvement of urban transportation in Greater Jakarta. The model can be employed and replicated to answer several issues in other agglomerations in Indonesia, such as Greater Bandung, Greater Bali, Greater Surabaya, Greater Medan, and Greater Makassar. Show more
Permanent link
https://doi.org/10.3929/ethz-b-000473046Publication status
publishedExternal links
Search print copy at ETH Library
Contributors
Examiner: Axhausen, Kay W.
Examiner: Mohammadian, Abolfazl
Examiner: Belgiawan, Prawira Fajarindra
Publisher
ETH ZurichSubject
Agent based modelling; Greater Jakarta; Urban air mobility; On-demand transport; Indonesia; Choice modelingOrganisational unit
03521 - Axhausen, Kay W. (emeritus) / Axhausen, Kay W. (emeritus)
02655 - Netzwerk Stadt u. Landschaft ARCH u BAUG / Network City and Landscape ARCH and BAUG
More
Show all metadata
ETH Bibliography
yes
Altmetrics