ANN ARBOR, MICHIGAN

In 1996, a test was conducted in Ann Arbor, Michigan that investigated bus priority using green extensions, red truncation (with and without compensation), and skip phasing (without compensation) at 14 intersections. The test showed that TSP could reduce bus journey times by an average of 6 percent (75 second reduction out of 22 minutes 41 seconds). These results were deemed inadequate since transit vehicle benefits were not sufficient to counter the additional delays incurred by non-transit vehicles.

CHINA

Bus priority strategies in China have shown more positive effects when TSP systems can be custom tailored to meet the individual needs of a specific area. In 1998, a small-scale test showed that a TSP system installed on a simple two-intersection arterial can reduce average bus journey times by up to 10.3 percent. Unfortunately, the study did not consider non-transit vehicles or the overall impacts on traffic operations.

In the United States; however, cooperative efforts between various transit and highway agencies have shown that highly effective TSP strategies can increase the number of people being transported as well as the number of cars being moved.

LOS ANGELES, CALIFORNIA

In the City of Los Angeles, a $10 million transit signal priority system was deployed on two corridors (Ventura Boulevard and Santa Monica-Beverly Hills-Montebello Route) based on TSP strategies originally developed in Curitiba, Brazil. The LA system was able to reduce bus journey times by 22 to 27 percent.

The system included 210 TSP equipped intersections, 331 loop detectors, and more than 150 on-board transponders. Each intersection was equipped with a controller cabinet, sensors, and inductive loops installed in each direction. Buses equipped with transponders were able to transmit bus ID into a central computer that was charged with controlling priority and keeping buses on schedule. At each intersection the computer granted early greens, extensions, free holds (holds a signal green until the bus passes through the intersection), and phase calls (brings up the selected phase that is not normally activated, i.e., left turn) as necessary. Buses were instructed to slow down or speed up to avoid interference with other buses en-route.

The TSP project used system software to reduce the amount of delay incurred by non-transit traffic. This was achieved through placing limits on green extensions if multiple buses successively requested priority at a common intersection. Project experts estimated that 30 to 40 percent of a 25 percent reduction in travel time realized on each route was the result of bus priority technology. The remaining benefits were the result of reducing the number of stops from one stop every 0.2 miles, to one stop every 0.8 miles.

PORTLAND, OREGON

In 1993 and 1994 bus priority systems were tested on Powell Boulevard and Multnomah Boulevard in Portland. The results showed that manually requested bus priority is inefficient since bus drivers often fail to call for priority. In contrast, automated signal priority (such as an infrared system) offers great potential. However, the author noted that additional testing is needed to accurately confirm the benefits of such an automated system.

SEATTLE, WASHINGTON

In the spring of 2000, the King County DOT implemented a 2.1 mile signal priority system on Rainier Avenue. The system used in-vehicle radio frequency transponders, roadside antennas, and transit priority request generators. The system was designed to control signal timings according to the following logic:

• Approaching priority vehicles were granted extended green intervals and shortened red displays; however, signal coordination between intersections had to be maintained, and the phase cycles for adjacent traffic could not be skipped.

Major findings of the study showed intersection delay was reduced by an average of 13 percent (2.3 seconds per vehicle) during the AM peak period. In addition, minor traffic movements on side streets, and in left turn lanes off the main street showed intersection delay reductions of 3 percent (0.9 seconds per vehicle). During the midday peak, intersection delay increased slightly; however, level-of-service remained the same.

In general, before and after studies conducted by King County Metro indicated the following:

• TSP reduced average bus delay by approximately 5 seconds per TSP equipped intersection.
• TSP reduced signal related stops by 50 percent for TSP equipped buses.
• TSP reduced bus travel time variability by 35 percent.
• Side street delay was likely unnoticeable to adjacent traffic during TSP operations since no cycle failures were recorded, and no one had to wait through more than one green cycle.
• Transit operator workloads were reduced as a result of fewer signal related stops.