This dataset is based on the ACFR Five Roundabouts Dataset. The original tracking data of over 23,000 traffic vehicles have been processed with an optimization-based filtering method to combat measurement noise and errors. Smooth velocity and acceleration signals are reconstructed. The processed recordings have then undergone a selection process using DBSCAN to remove the erroneous samples. The remaining samples contained in this dataset are considered representative of how average human drivers approach a roundabout scenario in daily driving.


We sincerely hope that sharing this dataset would help researchers in the relevant fields. Explanations on the dataset structure can be found in the README in the zip file. In case of any questions, please contact the authors via: Thank you!


The details of the processing method are presented in:

Y. Zheng, B. Shyrokau and T. Keviczky, "Comfort and Time Efficiency: A Roundabout Case Study," 2021 IEEE International Conference on Intelligent Transportation Systems (ITSC), 2021

(To be available online in Oct. 2021. Citation information will be updated.)


When using this dataset, please kindly cite our work above as well as the following:

A. Zyner, S. Worrall and E. M. Nebot, "ACFR Five Roundabouts Dataset: Naturalistic Driving at Unsignalized Intersections," in IEEE Intelligent Transportation Systems Magazine, vol. 11, no. 4, pp. 8-18, winter 2019, DOI: 10.1109/MITS.2019.2907676.


The original ACFR Five Roundabouts Dataset can be found via:


Extensive experimental measurement campaigns of more than 30,000 data points of end-to-end latency measurements for the following network architecture schemes is available:

  • Unlicensed IoT (standalone LoRa)
  • Cellular IoT (standalone LTE-M)
  • Concatenated IoT (LoRa interfaced with LTE-M)

Download to access all relevant files for the open data measurements.

Related Paper:


Exit Advance Guide Signs play an important role in driving safety on highway tunnels. With the increase of mountain highway in recent 10 years, how to set exit advance guide signs inside the mountain highway tunnels become a new problem in China. The paper highlighted the influence of exit advance guide signs on the driver’s eye movement in the mountainous highway tunnel.


The dataset contains high bandwidth voltage and current measurements of the main inverter of an electric vehicle. They were acquired from a Mercedes-Benz E-Vito on a testing ground in many different Operation Points (OP) listed in the following table:


The dataset consists of 14 hdf5 files containing the measured data. In addition, there are two python examples on how to handle the data and plot same results, and one readme file.
The hdf5 dataset can be accessed with many different tools like matlab, octave or python. If you want to use the python example, you must place the python-file and the dataset in the same folder. A recent version of python (it was tested with Python 3.9.2) with the following packages is needed: h5py; matplotlib; numpy; random; os; sys and scipy.

Python demo:
There are two python example demos to read and plot the hdf5 datasets included:
The first one reads a single operation point and plots the data in the time and frequency domain. (
The second one reads one dataset and calculates the short time Fourier transformation of all operations points in the dataset and plots a spectrogram. (
The demo is made as an example on how to handle the data and can be used for further analysis.
The datasets can also be accessed via matlab/octave, but therefore I refer to the online support.
If you have any further question about the dataset, please contact the author.


The proposed dataset, termed PC-Urban (Urban Point Cloud), is captured with an Ouster LiDAR sensor with 64 channels. The sensor is installed on an SUV that drives through the downtown of Perth, Western Australia (WA), Australia. The dataset comprises over 4.3 billion points captured for 66K sensor frames. The labelled data is organized as registered and raw point cloud frames, where the former has a different number of registered consecutive frames. We provide 25 class labels in the dataset covering 23 million points and 5K instances.


This dataset contains road networks used in experiments for DRL-Router, including Sioux Falls, Anaheim, Winnipeg and Barcelona.


The dataset collects the results of a survey of 325 respondents. Each respondent is asked to design a route from an origin to a destination taking into account the following considerations:

  • The route should avoid crowds to avoid getting COVID-19.
  • They should take into account the context provided: day, time, month, holiday period.

A total of 10 scenarios located in the city of Ciudad Real were designed.


Another raw ADS-B signal dataset with labels, the dataset is captured using a BladeRF2 SDR receiver @ 1090MHz with a sample rate of 10MHz


This dataset represents a subnetwork of public transportation in the city of Johannesburg. It contains counting of bus occupation of three significant routes as well as GPS location of Bus stations.


The data can be freely downloaded. It is made available for academic purposes


Both passenger demand and service supply are among the most important factors that determine the performance of urban rail transit system. It is not easy to find out optimal solution for the match between the passenger demand and service supply with traditional methods, due to the complexity of the combinatorial intelligent supply — demand matching problem. In order to get the comprehensively optimal matching degree, this paper transforms the multi-criteria problem into the distributed artificial intelligence optimization by using multi-agent dynamic interaction technique.