Lane-level localization and map matching for advanced CAV applications

This data repository is for the "Lane-Level Localization and Map Matching for Advanced Connected and Automated Vehicle (CAV) Applications" project. This project investigated and demonstrated the utility of lane-level map-matching and localization. In this project, data were collected in support of three folders for Tasks 2, 5, and 6.

Task 2: Lane-Level Mapping. Experimental data were acquired to assess the accuracy of the USDOT Mapping Tool. The data analysis used 39 feature points within about 200 meters of the intersection verified point and 55 feature points distributed over longer distances from the verified point (94 points total). Along with the data files, the repository includes a README file and two Matlab scripts that process the data.

Task 5: Demonstration. Experimental data were acquired to assess the probability of correct lane determination. Three road tests were performed. The data for each test is organized into its own subdirectory. The main directory contains a README file that discusses the file contents and how to process them using the included Matlab scripts.

Task 6: Simulation Study. Each simulation run created 4 .csv files: Chicago Intersection Queue information, Cranford Intersection Queue information, Iowa Intersection Queue information, and general vehicle information. Queue information consisted of the estimated queue information and actual queue information for each lane versus time. General vehicle information consisted of simulation time, vehicle id, vehicle speed, vehicle position, perturbed vehicle position, and vehicle direction. Each .csv file has column headers for distinction. In total there were 1200 .csv files: 4 .csv files for each simulation, 10 simulations for each scenario, and for the 30 scenarios described in the Simulation Scenario Section.