This git repository is linked to the research papers entitled "Building a Motorway Digital Twin in SUMO: Real-Time Simulation of Continuous Data Stream from Traffic Counters", and "A digital twin in transportation: Real-time synergy of traffic data streams and simulation for virtualizing motorway dynamics"
This repository includes all the files required to run the simulations used to produce the results presented in the papers:
(SCENARIO-I) Example of real-time simulation of actual traffic flow data (with minute resolution) obtained directly from Geneva motorway via ODPMS:
Notes: one needs to register at Open data platform mobility Switzerland (ODPMS) https://opentransportdata.swiss/en/ to get access to actual traffic data from traffic counters. Also, calculations in code are dependent on the step_length parameter, so it needs to be coupled with the parameter step length in .sumocfg to have everything working properly. Use the values for step length recommended at (https://sumo.dlr.de/userdoc/Car-Following-Models/EIDM.html). As well as, so far the EIDM model was not tested with option --step-method.ballistic.
- SUMO Geneva motorway network model and scenario files,
- instructions to register on the ODPMS platform and get access to actual traffic data,
- instructions to setup SQL database on a local computer,
- python code for:
(I) runtime access to Federal Roads Office (FEDRO) server via ODPMS (collecting actual data and storing it in real-time in a local SQL database),
(II) runtime traffic demand generation and online simulation calibration using SUMO's calibrators and actual traffic data as inputs to the running simulation scenario (ODPMS <-> SQL <-> python <-> TraCI <-> SUMO),
(III) the output excel files with results (comparison between actual traffic flows from Geneva motorway and the runtime simulated traffic flows by Digital Twin).
(SCENARIO-II) Example of dynamic flow generation in SUMO using external traffic flow data inputs with minute resolution:
Note: one can change the provided real workday traffic volume in workday_traffic.xlsx to see the effect of dynamic volume generation by SUMO's calibrator objects and foundation of DT-GM framework. Unrealistic traffic volume may cause instability in numerical computation in the flow model used by DFC mechanism
- SUMO Geneva motorway network model and scenario files,
- workday dataset (real traffic data with minute resolution),
- python code for:
(I) simulation of runtime demand generation (calibration) using external inputs traffic data,
(II) the output files of results (comparison of real traffic and the runtime simulated traffic).
Please use the following citation when referencing our work. Merci!
@article{KUSIC2023101858, author = {Krešimir Kušić and René Schumann and Edouard Ivanjko}, title = {A digital twin in transportation: Real-time synergy of traffic data streams and simulation for virtualizing motorway dynamics}, journal = {Advanced Engineering Informatics}, volume = {55}, year = {2023}, doi = {https://doi.org/10.1016/j.aei.2022.101858} }
@inproceedings{Kusic2022_DT_GM_ELMAR, author={Kušić, Krešimir and Schumann, René and Ivanjko, Edouard}, booktitle={Proc. of 64th International Symposium ELMAR}, title={{Building a Motorway Digital Twin in SUMO: Real-Time Simulation of Continuous Data Stream from Traffic Counters}}, year={2022} }
Note: current DT-GM framework uses SQL between ODPMS (traffic counters) and SUMO to permanently store actual traffic data received from traffic counters since we use data in the latter analysis
(I) Install Python (we tested DT-GM in version 3.9.13)
(II) Open cmd and navigate to the directory where you want to create Python virtual environment (can be a path to our folder you downloaded) and run commands (copy) in cmd to create Python virtual environment using terminal command (Note: if you have multiple pythons installed specify a version):
create virtual environment
python -m venv venvDTGM
activate virtual environment
venvDTGM\Scripts\activate
update pip
python -m pip install --upgrade pip
install necessary libraries from requirements.txt file
pip install -r requirements.txt
Instal SUMO (we tested DT_GM in version Latest Development Version May 30 2022 23:15:46 UTC), however, we highly recommend using the latest release of SUMO
(I) Install Microsoft SQL (we tested DT-GM in version: SQL Server 2019 Express and SQL Server Management Studio (SSMS) 18.11.1)
(II) run query below in SQL:
Execute:
Create Database SWISS_traffic_data
Then this
Use [SWISS_traffic_data]
Go
Then create table by executing this block
Create Table tblDetectors
(
ID int NOT NULL IDENTITY(1,1),
DetectorID nvarchar(50) NULL,
TimeStampUTC DATETIME NULL,
CarFlow int NULL,
CarSpeed float NULL,
TruckFlow int NULL,
TruckSpeed float NULL,
UnknownClassFlow int NULL,
UnknownClassSpeed float NULL,
)
Note: register at ODPMS https://opentransportdata.swiss/en/
and insert the token you were provided by ODPMS in the python script ODPMS_traffic_counters_request_live.py (see comments in the script).
Run the ODPMS_traffic_counters_request_live.py using terminal command:
activate virtual environemt
C:\Users\kkusic\Desktop\DT-GM>venvDTGM\Scripts\activate
navigate to folder SCENARIO-I
(venvDTGM) C:\Users\kkusic\Desktop\DT-GM>cd SCENARIO-I
run code for real-time traffic data collection
(venvDTGM) C:\Users\kkusic\Desktop\DT-GM\SCENARIO-I>python ODPMS_traffic_counters_request_live.py
For (SCENARIO-I): run live simulation (actual traffic from motorway traffic counters - ODPMS) <-> SQL <-> Python <-> TraCI <-> SUMO
Run python script <DT_GM_live24h.py> using terminal command:
activate virtual environment
C:\Users\kkusic\Desktop\DT-GM>venvDTGM\Scripts\activate
navigate to folder SCENARIO-I
(venvDTGM) C:\Users\kkusic\Desktop\DT-GM>cd SCENARIO-I
run On-The-Fly synchronized Digital Twin Simulation of Geneva motorway (DT-GM) (SCENARIO-II)
(venvDTGM) C:\Users\kkusic\Desktop\DT-GM\SCENARIO-I>python DT_GM_live24h.py
For (SCENARIO-II): run small example of dynamic flow generation by sampling external flow inputs (one workday traffic dataset from input_flow.xlsx) into the running simulation
activate virtual environment
C:\Users\kkusic\Desktop\DT-GM>venvDTGM\Scripts\activate
navigate to folder SCENARIO-II
(venvDTGM) C:\Users\kkusic\Desktop\DT-GM>cd SCENARIO-II
run dynamic flow simulation script:
(venvDTGM) C:\Users\kkusic\Desktop\DT-GM\SCENARIO-II>python DT_GM_workday_dynamic_flow.py
- Author Paulo Ribeiro, Master-Thesis: A Modelling Methodology Towards Automated Generation of Road Network Digital Twins, GitHub: https://github.com/paulinho-16, Master in Informatics and Computing Engineering at FEUP (Faculty of Engineering of the University of Porto)
If you use DT-GM or its tools to create a new one, we would be glad to add you to the list. You can send an email with your name and affiliation to kresimir.kusic@unizg.fpz.hr
As well, feel free to contact us if you experience any problems, so we can improve or add additional comments to make DT-GM easier to use.