Origin and Destination Study
The Association of Monterey Bay Area Governments (AMBAG) directed the Monterey Bay Origin Destination Study to gather information on the travel behavior of people who make interregional trips within the AMBAG region. They use this data to supplement the California Household Travel Survey in order to calibrate the AMBAG regional travel demand model. We evaluated a wide range of various innovative data collection techniques and chose a combination of four methods allowing us to provide an unprecedented view of travel behavior in the AMBAG region.
Combining Traditional Data Collection Methods with Big Data
Since interregional travel cannot be collected by a typical traffic count process, the standard approach is to gather data through the use of a survey. However, traditional survey methods tend to be costly and generally provide very small sample sizes. Vehicle classification counts play a pivotal role in any data collection survey as they can be used as a control total to refine the sample of interregional travel data. Infrared video cameras and computer software captured license plate data over a 24-hour period and transcribed the individual license plates. The California Department of Motor Vehicles then provided mailing addresses for our list of unique license plate numbers. We then mailed a survey to obtain information about the origins and destinations of the observed interregional vehicle trips, as well as information regarding trip purpose, trip frequency and demographic characteristics.
To supplement and compliment these data collection methods, we used Big Data in the form of both American Community Survey Census data and cell phone data obtained from AirSage. The Census data enabled us to fill in the missing demographic characteristics of the drivers. The cell phone data provided a very large sample of observed origin-destination for all types of travel within the AMBAG region in a format suitable for comparison with the local travel demand model. This made more sense than using Bluetooth technology, because this type of data collection would only provide intercept to intercept data, which does not provide the true origin or destination of the trip. The cell data also relies less on observed field data and user responses which can potentially introduce error and has a much lower cost per sample than traditional survey data or trip data collected using Bluetooth technology.
Developed state-of-the art methods to display the 1.4 million origin-destination records.
Origin-destination data is typically provided in very large and cumbersome trip tables that are difficult to visualize and draw conclusions from.
We used two types of heat maps to display the data:
Heat maps, which provide the intensity of trip generation in specific zones.
Desire line maps, which display the direction and magnitude of trip making.
By combining traditional data collection methods with the use of Big Data, AMBAG was able to obtain a massive amount of observed trip data to more thoroughly refine and calibrate their travel demand model.
The accuracy and additional data collected with the new license plate video technology alone increases the value of the survey. The ability to also collect cell phone data over 30 days and combine with Census data to infer the household demographics has great potential but has not yet been done. Combining both approaches on the same study is very exciting. We are looking forward to the advances the project can bring to data collection, travel modeling and transportation planning.Anais Schenk