README
(C) Raphael E. Stern, Shumo Cui, Maria Laura Delle Monache, Rahul Bhadani, Matt Bunting, Miles Churchill, Nathaniel Hamilton, R'mani Haulcy, Hannah Pohlmann, Fangyu Wu, Benedetto Piccoli, Benjamin Seibold, Jonathan Sprinkle, Daniel B. Work, December, 2017.

This file contains instructions for extracting data from the raw data files in the /data directory, as well as creating the figures and tables shown in “Dissipation of stop-and-go waves via control of autonomous vehicles: Field experiments”.

This data is published under the CC0 license and you are free to use the data as you wish, we only kindly ask you to consider citing the following work if you plan to publish subsequent results using the dataset:

Raphael E. Stern, Shumo Cui, Maria Laura Delle Monache, Rahul Bhadani, Matt Bunting, Miles Churchill, Nathaniel Hamilton, R'mani Haulcy, Hannah Pohlmann, Fangyu Wu, Benedetto Piccoli, Benjamin Seibold, Jonathan Sprinkle, Daniel B. Work, “Dissipation of stop-and-go waves via control of autonomous vehicles: Field experiments.” Transportation Research Part C: Emerging Technologies (in press).

===================================================
Directories
===================================================

There are two directories: code and data.

-code: all the code necessary to extract and synchronize the data and make plots
-data: all raw data presented “Dissipation of stop-and-go waves via control of autonomous vehicles: Field experiments”

====================================================
Using data without MATLAB
====================================================

All data provided, with the exception of the log files recorded on the CAT Vehicle, are provided in CSV format. Within each experiment data directory (Exp_A, Exp_B, and Exp_C) are four .csv files, one .mat file and a directory titled “OBD”. They are described as follows:

accelerations_smoothed.csv — smoothed accelerations in m/s/s. Each column corresponds to a vehicle (in the same order as in Table 5 in the paper. Note, however, that vehicle number 22 is only used in Experiment C). The time step for each row of the CSV corresponds to 1/30 s

velocities_smoothed.csv — smoothed velocities in m/s. Each column corresponds to a vehicle (in the same order as in Table 5 in the paper. Note, however, that vehicle number 22 is only used in Experiment C). The time step for each row of the CSV corresponds to 1/30 s

displacements_smoothed.csv — smoothed position of each vehicle in m. Each column corresponds to a vehicle (in the same order as in Table 5 in the paper. Note, however, that vehicle number 22 is only used in Experiment C). The time step for each row of the CSV corresponds to 1/30 s

displacements_unsmoothed.csv — raw vehicle positions extracted from the video footage in units of m. Each column corresponds to a vehicle (in the same order as in Table 5 in the paper. Note, however, that vehicle number 22 is only used in Experiment C). The time step for each row of the CSV corresponds to 1/30 s.

exp_[Exp. letter]_CATVehicle.mat — log files from the CAT Vehicle used to synch vehicle trajectories with each vehicle’s OBD-II file if using the scripts above

OBD/veh[num]_fuel_data.csv — csv for each vehicle’s OBD-II data. The columns (with units) are: [time (s), fuel economy (l/km), fuel rate (l/hr), OBD-II velocity (km/hr)

===================================================
Extracting data
===================================================

Run the synchronize_test_data.m script to extract the raw ROS data from the CATVehicle, video data, and OBD-II data for each vehicle. This will save a file ‘synchronized_data.mat’ with a structure for each experiment conducted. These structures contain all the data and metadata about each experiment. This will also synchronize the times from the three different measurement devices. All scripts have been written using MATLAB 2017a.

====================================================
Plotting data
====================================================

To plot the data, run the plot_synchronized_data.m script (after extracting and synchronizing the data). This will create all the figures and tables in the paper.

====================================================
Description of all scripts and functions
====================================================

plots.m — function used to make all the plots for each experiment

plot_velocities.m — function used to plot the velocities and velocity standard deviations for a given experiment

plot_trajectories.m — function to plot the vehicle trajectories, the CAT Vehicle’s trajectory is plotted in red

plot_std_per_time.m — function used to plot the velocity standard deviation at each instant in time to determine when a traffic oscillation has arisen

plot_kappas.m — function to plot extreme breaking events

plot_fuel_consumption.m — function to compute and plot the fuel consumption for each period of the experiment

plot_events — function to label graphs with the start time of each period of the experiment

plot_all_std.m — function to plot the standard deviation of all vehicles per time period on the velocity plot

get_std_for_span.m — function to compute the velocity standard deviation of all vehicles over a particular time span

get_std_for_signal_span.m — function to compute the standard deviation of an arbitrary signal over a time span

get_means_for_span.m — function to compute the mean velocity of all vehicles over a specific time span

generate_experiment_table_single.m — function to output all computed values in a table to summarize a single experiment

generate_experiment_comparison_table.m — function to output all computed values in a table to compare different experiments

elapsed_time.m — compute the time that has elapsed since a given reference time
