Lead Presenter Biography

Parag Gupta, a data science enthusiast, and seasoned travel demand modeler with 7 years of experience. With a passion for leveraging data to drive strategic decision-making, Parag Gupta has developed expertise in creating and implementing predictive models to forecast travel demand for various modes of transportation. Over the years, Parag Gupta has honed their skills in statistical analysis, data visualization, and data communication. He has successfully used these tools to identify trends and patterns in travel data. In their most recent role, Parag Gupta was instrumental in developing SERPM 9 External Travel Demand model, Congestion Dashboard, Desktop Application for RSLAM (Richmond Simplified Land Use Allocation Model) and various post processing tools for TDM which has improved decision-making and enhanced the overall user experience. With a strong work ethic, Parag Gupta is committed to continuously improving their skills and staying up to date with the latest industry advancements

Co-Authors

Presentation Description

Richmond Simplified Land Use Allocation Model (RSLAM) was developed to support the modeling and planning effort, and to facilitate the region’s scenario planning process. It is a data-driven growth allocation model that creates a zonal data file suitable for input to and use by the RRTPO regional travel demand model. The model begins with a zonal data file from the base year, and regional future year control totals for each zonal data item, and allocates the growth specified in the control totals to individual zones, thereby creating a future year zonal data file. It handles the eight jurisdictions in the region separately.
It uses local data to guide the way that growth is allocated. It is a fast R application and has been calibrated using local data to replicate current forecasts. Besides the future year zonal data file, the model produces tables and maps to illustrate model assumptions and results and is controlled by a Graphical User Interface (GUI).
The GUI is a modern web-based user interface tool developed in R programming language, that is an open-source software environment for statistical computing and graphics. It takes advantage of the state-of-the-art Shiny package that makes it easy to build interactive web-apps straight from R. These applications can be hosted as standalone applications on webpage or embed them in R to use them as standalone desktop applications.
The final application designed gives users a simple and information-rich experience in modifying the inputs, in reading, and visualizing the outputs. Users can directly view the modified files in the GUI and write new scenario-specific files. The GUI offers full control on creating scenarios, editing the control totals, changing densities, and modifying Travel Analysis Zone’s (TAZs) attributes. Users can name their scenarios and select input files to be modified. Once the file is selected, the GUI dynamically renders the selection fields/columns to be changed. This helps envisioning the scenarios to be modeled. Once the user completes modifying the file, they can finalize the scenario by writing the files to the system.
Once the files are written, users can run RSLAM directly from the user-interface just by pressing ‘Run Model’ button. It takes less than two minutes to allocate the desired growth and produce outputs for the scenario. Users can read and visualize the outputs directly from the GUI itself. After the model run is complete, users can navigate to the output tab. The output tab is divided into three sections that illustrate the results in various maps and chart format.
Overall, the experience designed is user-friendly and offers a seamless integration to the scenario planning and travel demand modeling toolkit. The procedure was developed to be run locally or on a webpage using open-source R-shiny platform. This is one of the very few innovative examples that can run a model from a webpage. 

Presentation File