Optimizing bikeshare service to connect affordable housing units with transit services
Data files
May 30, 2023 version files 965.14 MB
-
ca_od_main_jt00_2019.csv
-
README.md
Abstract
This research studies the potential of bikeshare services to bridge the gap between Affordable Housing Communities (AHC) and transit services to improve transport accessibility for the residents. In doing so, the study develops an agent-based simulation optimization modeling (ABM) framework for the optimal design of the bikesharing station network considering improving accessibility as the objective. The study discusses measures of accessibility and uses travel times in a multi-modal network. Focusing on the city of Sacramento, CA, the study gathered information related to affordable housing communities, detailed transit services, demographic information, and other relevant data. This ABM framework is used to run three stages of travel demand modeling: trip generation, trip distribution, and mode split to find the travel time differences under the availability of new bikesharing stations. The model is solved with a genetic algorithm approach. The results of the optimization and ABM- based simulation indicate the share of bike and bike & transit trips in the network under different scenarios. Key results indicate that about 60% of the AHCs are within 25-minute active travel time when the number of stations ranges from 25 to 75, and when the number of stations is increased to 100, most AHCs are within 40 mins of active mode distance and all of them are less than an hour away. In terms of accessibility, for example, having a larger network of stations (e.g., 100) increases by 70% the number of Points of Interest (for work, health, recreation, and other) within a 30-minute travel time. This report then provides some general recommendations for the planning of the bikesharing network considering information about destination choices as well as highlighting the past and current challenges in housing and transit planning.
Methods
The dataset for this study was collected from various sources including Affordable Housing Communities Data, demographic information from the American Community Survey, OpenStreetMap road network, Points of Interest (POIs) data, General Transit Feed Specification (GTFS) data, and Longitudinal Employer-Household Dynamics (LEHD) data. You will need to download all the data from their sources except for the LEHD data.
The data was collected from the Sacramento Housing and Redevelopment Agency and the office of the State Treasurer which maintains a list of affordable housing projects. The data was then processed to identify and eliminate duplicates and old projects that were scrapped. The final list consisted of 149 affordable housing projects spread across the city. Demographic information like household income, number of family and non-family households, and number of occupied and vacant housing units were sourced from the 5-year estimates of 2020 American Community Survey data. The LEHD Origin-Destination Employment Statistics or LODES data were used to identify origin-destination matrix with census blocks with residences/homes as the origin and the census blocks with workplaces/ offices as the destination. Bike and Walk Network Data from OpenStreetMap and POIs Data from OpenStreetMap were also used in the study.
Usage notes
The data is sourced from several data sources such as the Sacramento Housing and Redevelopment Agency, American Community Survey, OpenStreetMap, General Transit Feed Specification (GTFS) data, and Longitudinal Employer-Household Dynamics (LEHD) data. These data sources provide data in various formats such as CSV, XML, JSON and shapefile. You will need to download the data from other sources before using the code.
For coding, you can use python and its libraries such as pandas, numpy, geopandas, and osmnx to handle and process the data. Python is an open-source programming language that can be run on multiple platforms such as Windows, Mac, and Linux. The Jupyter Notebook is an open-source web application that allows you to create and share documents that contain live code, equations, visualizations, and narrative text.
If the data files are in proprietary format, you can use open-source libraries and software to convert them to a more common format, such as csv, which can be opened with a basic text editor or spreadsheet program, such as Microsoft Excel or Google Sheets.