Keep Your Buses Running On Time

In this era of promoting alternative transportation modes, traffic signal professionals need to develop and implement methods to provide improved transit operation. The city of Portland and the Tri-County Metropolitan Transportation District of Oregon (Tri-Met) have a long history of joint efforts for improving transit operations, including traffic signal priority.

Portland’s history of signal priority for transit started with the light rail system, which began revenue service in 1986 and had emergency vehicle-like pre-emption at many of the traffic signals in Portland. The level of priority was actually increased at 11 signals approximately two years after initial startup.

Prior bus priority experience in Portland has been less extensive. Two separate bus priority tests have been conducted in the city–the Powell Boulevard test in 1993 and the Multnomah Boulevard test in 1994.

These projects included evaluation of the technology to transmit priority requests from the bus to the traffic signal controller and tests of priority algorithms. The transmission technologies have included the TOTE System by McCain Traffic Supply Inc. of Vista, Calif., the LoopComm System by U.S. Traffic Corp. of Sante Fe Springs, Calif., and the Opticom Priority Control System by 3M of St. Paul, Minn.

The signal priority algorithms included the green extension/red truncation method and queue jumps. An additional test of the Opticom system with green extension/red truncation control was implemented by the Oregon DOT in the west suburban Portland region in a 1996 T.V. Highway Study. All tests were done at traffic signals with Type 170 controllers and Wapiti IKS software, manufactured by Wapiti Micro Systems Corp. of Prineville, Ore., the standard throughout the Portland region.

Some of the basic conclusions and lessons learned from these tests included:

• The method of transmitting a priority request from the bus to the traffic signal should not require bus driver intervention;
• The Opticom system worked reasonably well in signaling the presence of the bus to the traffic signal. One advantage of selecting Opticom was that many traffic signals in the Portland region already have Opticom receivers installed for emergency vehicle operations, which should make future expansion of bus priority much easier to accomplish;
• The green extension/red truncation algorithm in the Wapiti IKS Type 170 software was limited in its flexibility. While a signal could be kept in coordination, the left-turn phase lengths had to all be driven to minimum values, which was unacceptable for the major intersections included in the Powell Boulevard test;
• Detector placement at intersections is an important factor in the proper operation of bus priority measures–particularly for near-side stops; and
• Tests found that traffic signal priority devices do provide a net reduction in round-trip bus travel times. Overall, benefits may include improved fuel costs, less equipment wear and lower operator fatigue and stress.

In addition to these tests the city, in conjunction with Tri-Met, conducted the "Transit Preferential Streets Study."

The purpose of this study was to:

• Develop a tool box of potential bus priority techniques. These techniques included other non-signal related concepts such as queue bypass lanes, bus stop relocations, curb extensions, etc.; and
• Analyze five major transit corridors in Portland and recommend a range of treatments to consider for providing bus priority.

This study provided a planning framework to determine where and how bus priority, not just bus signal priority, would be provided in Portland. Both the city and Tri-Met have agreed to embark on an aggressive overall program to provide bus priority on the most congested corridors in Portland.

The goals for transit priority should be to reduce bus travel times and to improve schedule reliability, hopefully leading to increased bus ridership.

The techniques identified in the "Transit Preferential Streets Study" are being aggressively applied to the highest priority corridors in Portland. Bus signal priority is only one part of providing preferential treatment for transit, and the reader needs to remember to address all aspects for a successful program.

As part of the TEA-21 High Priority Project category, the city of Portland will receive an estimated $4 million in federal funds to provide signal priority for emergency and transit vehicles (with 20% local match the total project cost is approximately $5 million). This project will build upon two key existing elements–Tri-Met’s successful automated vehicle location (AVL) system and Portland’s optical emitter system for providing signal priority for fire vehicles.

Project elements include:

• Installation of optical emitters on Tri-Met’s entire standard bus fleet, 775 fixed route buses and revising the on-board AVL system logic to activate the emitter at the appropriate time and location; and
• Installation of intersection optical detectors at intersections without Opticom and installing an upgraded 170 controller or a "2070" generation controller at all intersections.

The upgraded 170 software will have an improved red truncation/ green extension algorithm. The new "2070" controller software will have additional bus priority algorithms. About 190 intersections will receive Opticom while a total of nearly 250 intersections will receive the new controllers. The project should provide complete signal priority for five major bus routes in Portland.

By equipping the entire bus fleet with emitters at this time, signal priority can eventually be extended to other corridors by merely updating the signal controllers, not only within Portland but at other suburban intersections as well.

The key element to the Portland bus priority concept is Tri-Met’s AVL system developed by Orbital Sciences Corp., Dulles, Va., which is called the Bus Dispatch System (BDS). The BDS is based on the "Smart Bus" concept where each bus determines its location within 16.5 ft using an on-board GPS receiver. The GPS receiver, as well as many other bus functions, is connected to the bus’ on-board computer, or central logic unit. The central logic unit has all route and schedule information and is able to determine if the bus is late or early. The bus reports itself early or late to the dispatch center–if it is outside of a variable window usually set at two minutes early or seven minutes late. For requesting signal priority, the initial late threshold has been set at 90 seconds, although this parameter can be changed later.

The project plan is to have the smart bus activate the Opticom emitter, or bus-to-wayside communication device, only when the bus determines that it is late.

As noted above, the early tests of Opticom were satisfactory. Combined with the existing installed base of Opticom intersection detectors, the city and Tri-Met have decided to use Opticom for communicating the request for priority from the smart bus to the smart intersection. The buses will use Opticom emitters with visible light screens so that other transportation system users don’t see a flashing strobe on the buses. Also, the buses will only receive low priority through the Opticom system. All emergency vehicles will have high priority that will always override any low priority operation.

In the future, we anticipate using the class number as an indicator of the level of priority, based on such things as lateness or route importance. These differing class or priority values can then be used by the traffic controller to determine which bus should get the higher priority when two conflicting routes cross each other.

As noted above, the bus’ on-board computer will initiate the request for priority any time the bus determines that it is falling behind schedule. Initially the emitter will be activated anytime the bus is late–90 seconds behind schedule–and is within the city limits for Portland. After a future system upgrade, each route will have specific "activation zones" noted in the route database.

When this upgrade is in place, the bus will activate the emitter only when the bus is in the activation zone and only when the bus is late. Also, to address near-side bus stop issues, the emitter will be deactivated anytime that the door is open.

As could be expected, Orbital and 3M are close partners with the city and Tri-Met in making this bus-Opticom interface a reality.

The final key element to the whole program is implementing an improved bus priority algorithm. The existing limited memory in the Type 170 controller with a Motorola 6802 chip has been a constraint in adding enhancements to the existing Wapiti IKS program. As a result, the city worked with Wapiti Micro Systems to develop an improved version of their IKS program that runs on the HC11 chip for the 170. Part of the software upgrade will include providing a more robust green extension/red truncation algorithm than exists in the current software. The overall key will be to provide several variables that provide the traffic engineer with a wide range of operating choice.

In the future, the city of Portland intends on using a 2070 level controller with still more improved software. Wapiti, in conjunction with the Northwest Wapiti Users Group, has defined improvements to all priority and pre-emption operations by developing a wide range of recovery options. These improved recovery options allow the local traffic controller to use a logic tree in determining how and when to transition back to normal operation.

For instance, the controller may allow those phases that were skipped or minimized during pre-emption or priority to receive additional time following a pre-emption or priority operation. The recovery tree also may allow a variable return phase based on where the controller was just prior to pre-emption or priority. This 2070 level software is still a year away from being a reality.

The pilot route, Line 4 Division/Fessenden, was activated in mid-July 2001. This route covers approximately 90 traffic signals. Sixty of the 90 signals now have bus signal priority. Twenty signals on the bus mall are not planned to include bus signal priority and the other 10 signals have other technical issues like light rail pre-emption that preclude bus signal priority at this time.

Initial results have been positive. Preliminary numbers include:

• Line 4 buses experienced an 8% improvement in afternoon peak on-time performance; and
• For trips where the buses have been later than 90 seconds, the aggregate run-time was reduced by 2.3 minutes for the peak hour/peak direction of travel.

These results point to an issue that we always knew would arise–buses can never go faster than the schedule. If we want to make true travel time reductions, then we need to reduce the schedule time. For the pilot line, that reduction occurred in early September. Up to 10 minutes have been reduced from some schedules, depending on previous on-time performance and anticipated signal priority benefits. No results from this new schedule operation were available prior to when this article went to print. More detailed analysis should be completed by mid-October, although we certainly expect that the buses will be able to meet their new schedule.

Realizing true bus signal priority benefits will be an iterative solution. We expect to continue reducing the schedule until bus priority is required nearly all of the time just to keep on schedule. That’s how we plan to maximize priority for buses in Portland.

The results from the pilot corridor will be further analyzed this fall to determine what worked best. Bus signal priority will then be extended to the remaining four corridors, which should be completed by spring 2002. At that time, the city should have about 250 signalized intersections with bus signal priority.

After this project is completed, the city and Tri-Met will work on analyzing other corridors in need of bus signal priority. Also, Tri-Met will use the results of the Portland project to work with other local jurisdictions and the Oregon DOT to extend bus signal priority beyond the city limits of Portland.

Tri-Met and the city of Portland are excited about this important project. We believe that this large-scale bus priority installation may well serve as a model for other cities considering similar investments in making bus transit work better.