Repeated Route Naturalistic Driving Dataset

Data Sample R2ND2 Paper

OUT NOW, SCROLL DOWN TO GET THE DOWNLOAD LINK

Bikram Adhikari

About R2ND2

Repeated Route Naturalistic Driving Dataset (R2ND2) is a dual-perspective dataset for driver behavior analysis constituent of vehicular data collected using task-specific CAN decoding sensors using OBD port and external sensors, and (b) gaze-measurements collected using industry-standard multi-camera gaze calibration and collection system. Our experiment is designed to consider the variability associated with driving experience that depends on the time of day and provides valuable insights into the correlation of these additional metrics on driver behavior. The data is collected among 10 participants with varying driving expertise. Each driver drive through highway and secondary route mapped in Route map section in morning, day and evening. The research is one of many efforts from Mason Living Innovation Lab towards safe driving practices and intelligent transportation systems. Meet our team:
Bikram Adhikari
Zoran Duric
Duminda Wijesekera
Bo Yu

Sensor Setup for the experiment

OBD Scanner Setup

We used the OpenDBC protocol to decode the CAN message using the following sensors:

  1. Arduino UNO.
  2. SparkFun CAN-BUS Shield.
  3. GPS Receiver EM-506 SparkFun.
  4. YOST Labs TSS-WTS-S v2.6a IMU.

Gaze-Detection Setup

We used the industry standard Gaze Detection and Measurement sensors from SmartEye constituent of the following:

  1. 3 x Smart Eye IR cameras (Dimension:17x31x31mm, Resolution: 3MP, Lens Focal Length: 16 mm).
  2. 2 x IR emitters (Wavelength: 850 nm).
  3. Logitech C922 HD Webcam (1080p).
  4. GNSS200L USB GNSS (Global navigation satellite system) Receiver.
  5. SmartEye pro adapter module for data sync.

Experimental Route Map

The experiment was carried out in Highway and Secondary Routes between George Mason Arlington Campus and George Mason Fairfax Campus, the specific route maps with the GPS speed are illustrated as:

Highway Route

Secondary Route

Data visualization

There are 224 data features to work with, 32 features from the OBD scanner and 192 from the eye gaze measurement system. Each features can be visualized and compared against time or any other metric of measurement

Vehicular Data visualization

Considering speed to be a common measure of interest, here we have visualized speed from three different participants in three different time of day while driving through Highway.

Gaze Measurements Visualization

We are excited to share that we are able to provide actual scene footage with blurred license plate and pedestrian faces. Along with that we also provided extracted gaze-only video as seen on the left below. The gaze-mapping can be used in multiple ways like through gaze heatmaps or line plots of gaze-progressions as seen in the right below.

Data Download



License Agreement

R2ND2 is publicly available to advance research in driver behavior analysis. The dataset consists of non-personally identifiable data features. However, if you discover any information that discloses and violates your privacy, we kindly request that you contact us, and the data will be promptly removed.

To download and use the data, you must agree to the following rules:
  1. The dataset is exclusively for academic and non-academic research purposes. It should not be used for any commercial purposes, including, but not limited to, selling, licensing, sublicensing, or any other activities aimed at monetary gain.
  2. The data must not be modified, adapted, reverse-engineered, decompiled, or disassembled, and it should not be distributed in any form other than its original format.
  3. Proper attribution and citation of the dataset are required in any publications, presentations, or reports. The BibTeX citation is provided as follows:
    [@ARTICLE{10818997, author={Adhikari, Bikram and Durić, Zoran and Wijesekera, Duminda and Yu, Bo}, journal={IEEE Transactions on Intelligent Transportation Systems}, title={Repeated Route Naturalistic Driver Behavior Analysis Using Motion and Gaze Measurements}, year={2024}, volume={}, number={}, pages={1-11}, keywords={Vehicles;Cameras;Driver behavior;Sensors;Decoding;Accuracy;Roads;Data mining;Calibration;Kinematics;Driver behavior analysis (DBA);naturalistic driving experiments (NDE);controller area network bus (CAN Bus);inertial measurement unit (IMU)}, doi={10.1109/TITS.2024.3520893}} ]
  4. While we have made efforts to ensure the accuracy and reliability of the dataset, we disclaim all liability for any errors, omissions, or inaccuracies in the dataset.
  5. We will not be liable for any direct, indirect, incidental, or consequential damages arising from the use or inability to use the dataset.
Reach out to badhika5@gmu.edu in case of any concern with the data or code for analysis.

and


Data Repository