While safety seems to be a significant factor when choosing to use these new modes, this model utilizes the notion of perceived safety to model travel behavior in inner urban areas. Therefore, the developed model is built on the hypothesis that perceived safety affects travel behavior of e-scooter users and is related to road environment. It combines ordinal logistic regression model, which predict perceived safety in different road environments under different traffic flow conditions, with discrete choice models which give the mode choice. At the same time, it creates a path to combine agent-based transport modeling with spatial analysis and GIS tools. The input parameter is the road network which consists of links and nodes.
Perceived safety per urban transport mode is calculated by the following equations:
Perceived safety level is integrated as an additional parameter in MATSim scoring function. This allows the use of the developed model in agent-based traffic simulations. MATSim simulates and scores many alternative travel plans (different mode or route?) in order to define an equilibrium point where agents cannot further improve their scores. Below, the new MATSim scoring function is presented (two modeling alternatives):
The beta parameters of the model equations have been estimated based on a survey which combines a rating experiment with a stated preferences experiment. Four different road environments are assessed in this survey, namely: type 1: urban road with sidewalk < 1.5 m wide, type 2: urban road with sidewalks ≥ 1.5 m wide, type 3: urban road with cycle lane and type 4: shared space.
The Perceived Safety Choices repository contains:
- survey_design
- raw_data: collected survey data per survey block
- psafe_models: it contains the data processing of perceived safety rating data + data analysis of perceived safety ratings in R using Rchoice package. The output of this analysis are the beta parameters per mode + figures are included in the folder.
- choice_models: it contains data processing of choice data + model development using PandasBiogeme. The output of this analysis is beta parameters of choice model
- datasets: datasets of perceived safety ratings, sociodemographic characteristics and mode choices. These datasets can be used in other road networks (no need for new data collection).
- network_analysis: using pyshp, shps of nodes and links, in a very specific data format (see network examples), are imported to estimate perceived safety per link. The user has to provide these shp and run the code. The output of this process are xml network file (lxml toolkit is used) capable for MATSim simualtions and csv file, which can be imported in GIS and joined with shp for mapping purposes.
- prediction_routing
- indicators
You can run all the steps of the Perceived Safety Choices model from Perceived_safety_choice_model.py. Analytical instructions are included there (with comments).
+++ The contribution to MATSim is under development. In essense, it is an updated version of Bicycle contribution following a more universal approach fully based on perceived safety parameter and covering all micro-mobility modes.
Papers:
- Tzouras, P. G., L. Mitropoulos, E. Stavropoulou, E. Antoniou, K. Koliou, C. Karolemeas, A. Karaloulis, K. Mitropoulos, M. Tarousi, E. I. Vlahogianni, and K. Kepaptsoglou. Agent-Based Models for Simulating e-Scooter Sharing Services: A Review and a Qualitative Assessment. International Journal of Transportation Science and Technology, 2022. https://doi.org/10.1016/j.ijtst.2022.02.001.
This model was developed for SIM4MTRAN project that aims to develop an innovative integrated decision support tool for the design of micro-mobility systems and services. The results will be used to create a guide for the design of micro-mobility systems in urban areas in Greece supporting policy making process.
This research project has been co-financed by the European Regional Development Fund of the European Union and Greek national funds, National Strategic Reference Framework 2014- 2020 (NSRF), through the Operational Program Competitiveness, Entrepreneurship, and Innovation, under the call RESEARCH – CREATE – INNOVATE (project code: T2EDK-02494 and name: SIM4MTRAN).