Applications of the BHL Data

Applications of BHL include microtraffic studies, simulation calibration and validation, and field-testing of detection equipment and other hardware.

The loop detector system collects individual vehicle actuations from all 164 loops in the BHL area every 60th of a second and archives both the activation data and a large set of derived data which includes vehicle stream data which can be used for headway studies, gap analysis, and merging studies. The vehicle lengths are also available allowing classification of the freeway traffic. The video data provides detailed ground-truth data on a segment that is 1/3-of-a-mile long. If an other kind of data is required, CCIT can work with researchers to help collect and process this information differently. Contact CCIT at bhl@calccit.org with any questions.

Some hypothetical user scenarios for using the raw and processed BHL data include:

Model Calibration

Say Jamie at UCI needs to analyze the effects of a new overpass in Irvine using the Paramics simulator. The network is already coded and available on the UCI website, but its driver model parameters need to be calibrated. Jamie goes to the BHL Data Repository on the web and retrieves the network data from the ATMS testbed database. He then retrieves some macroscopic data from the BHL 30-second loop database. Then he retrieves 30 minutes of BHL vehicle trajectory data. He finds a 30-minute period with the type of traffic density and speed that is similar to the situation he is trying to simulate. He downloads the data. Using the trajectories, Jamie calculates the inter-vehicle spacing, lane changing, and preferred speed parameters and uses them in the Paramics model. He makes more fine adjustments by calibrating the simulator against the macroscopic data from the same region.

Emissions modeling

Or perhaps Larry at UC Riverside is developing a model of vehicle emissions. He has conducted some experiments on vehicles at various speed and accelerations. Now he would like to calculate the emissions for a whole region based on macroscopic measurements from loop detectors. He needs to develop a model that relates vehicle trajectories to macroscopic measurements such as 30-second speed and flow. He retrieves 10-minute BHL vehicle trajectories and loop detector data for the same time periods. He uses these datasets to develop a relationship between 30-second loop measurements and vehicle acceleration profiles.