There are only limited opportunities to add significantly to the transportation infrastructure in the Puget Sound region in the next 15 to 20 years. For this reason, there is a growing interest in improving the efficiency of the existing system. One way to do this is to increase the average vehicle occupancy (AVO) on freeways and arterials. Programs to accomplish this must be capable of evaluation. Therefore, accurate and up-to-date information on AVO is required.
This research project investigated various methods to measure AVO in order to determine the feasibility and costs of a continuous, ongoing data collection program. Since it was determined that there are no promising approaches using automatic methods employing new technology, the study focused on the use of human observers. The degree of accuracy was studied using three observers counting the occupancy of the same vehicles at the same time. The results showed the observers can be highly accurate (correct 97 percent of the time) and that environmental conditions such as weather, light, traffic density, and traffic speed do not have exceptionally strong effects on accuracy (within reason). Furthermore, observers can easily count up to a half hour at a time without fatigue affecting their performance.
Taking the results of this and previous research into account, it was determined that it is possible to provide quarterly counts of AVO at 26 sites that are accurate to within about 1.5 percent for about $50,000 per year. This is about the cost of one data analyst, when benefits and overhead are taken into account.