Open Data

Map-Matched Trajectory Data with Signal Phase Data Collected by AACVTE

General Data Description

The real-time CV data are forwarded from the infrastructure side (i.e., RSU) through the IPv6 network. The SPaT data are directly collected from the signal controllers, and the BSM data containing trajectory segments when the vehicles are within the communication ranges of RSUs are collected by RSUs installed at intersections.

Since the raw SPaT and BSM data follows SAE J2735 standard and are encoded with the UPER rules, a message decoder is developed to transfer the hex strings into C++ class objects for further processing.

The BSM data collected by the RSUs is at 10 Hz frequency. After decoding, it contains vehicle ID, timestamps, and the GPS coordinates of the vehicle that can be used to construct the trajectory. The SPaT data sent from the signal controller includes a timestamp and states of all signal phases at the intersection at 10 Hz as well.

Data pre-processing: Map-matching and phase mapping

Map-matching is an important pre-processing step that connects BSM data with road geometry. Since the trajectory data are collected near intersections (Due to the communication range), we only need to match the trajectory to its corresponding movement.

After map-matching, however, the mapping relationship between the signal phases and vehicle movements (e.g., left-turn and through movement of four directions) are not known. So we grouped the trajectory data to the corresponding movements by applying signal phase mapping identification algorithm. More details of the algorithm can be found from the paper here.


Safety Pilot Model Deployment Data Collected by Additional Data Acquisition Systems (DAS)

This data were collected during the Safety Pilot Model Deployment (SPMD). The data sets that these entities will provide include basic safety messages (BSM), vehicle trajectories, and various driver-vehicle interaction data, as well as contextual data that describes the circumstances under which the Model Deployment data was collected. Large portion of the data contained in this environment is obtained from on board vehicle devices and roadside units. The raw data can be downloaded here.

MTL is actively processing the SPMD data collected by additional data acquisition systems (DAS) so that it could be utilized more conveniently by the researchers. Please follow us closely for the upcoming available open data.