UC Irvine

Traffic Monitoring is at the center of any Intelligent Transport System. Current traffic monitoring sensors are challenged to deliver in the evolving landscape of connected, autonomous and alternative fuel transportation systems. This dissertation explores the feasibility of emerging LiDAR technology for traffic monitoring, in an infrastructure-based, side-fire LiDAR configuration. LiDAR technology was investigated in terms of providing both the core data elements of existing traffic monitoring systems such as vehicle counts and speeds, as well as more high-resolution data elements required for future connected and autonomous vehicles such as relative positions of vehicles on a roadway, vehicle lateral and longitudinal positions within a lane, physical attributes of individual vehicles, and vehicle microscopic trajectories. LiDAR sensors were deployed at both dense urban corridors and rural highway locations. At the urban location, the LiDAR estimate of vehicle counts across lanes was between 87% to 110% of a baseline calibrated sensor’s vehicle counts. At the rural highway location, microscopic trajectories for vehicles were derived at 0.1 second resolution, enabling detection of anomalies in vehicle behavior. In addition, the precise lateral positions of heavy-duty vehicles were derived at the urban corridor location, with a particular interest being future safety assessment for loss of control of autonomous heavy-duty trucks. The high-resolution traffic data elements derived from this research can assist in detecting anomalous behavior of vehicles, whether from impaired driving or loss of effective autonomous control, with road safety assessments, and in providing inputs for microscopic road emission modeling.