Dept. of Transportation
Metro Transit Division

King Street Center
201 S Jackson St
Seattle, WA 98104
Metro Online Home


Objective 1: Examine the functions and technical approaches of the smart bus off-the-shelf onboard equipment in the King County Metro Transit environment.

Conclusion: Good technical performance and potential, given a short-term effort

  • The technical performance assessments of the demonstration equipment conducted for this evaluation produced a range of results, understandable given the short duration of the demonstration. Most participating staff were impressed at how well the equipment performed, given the demonstration was conducted in advance of a full requirements development, procurement and acceptance testing process.
  • Some technical performance issues identified by staff, for example, in the APC and TSP assessments, were regarded as fixable by additional effort that would be justified by a permanent installation.

Automatic Passenger Counters:

The demonstration APC equipment performed well overall. With no optimization or iterative "fine-tuning" the demonstration APC system met or exceeded 12 out of 18 current KC Metro APC performance criteria. However APC staff noted the criteria would be met if the effort were allowed more time to fine-tune the demonstration equipment.

Data Transmission to Transit Signal Priority (TSP) equipment:

Vehicle schedule deviation and passenger load data were transmitted and read by the signal priority system, demonstrating the real-time dynamic data from the onboard system could be transmitted to the TSP system.

    • Data were not received for 18 of the 52 runs, for undetermined reasons. The TSP tag design required that the value of the lateness field was passed to the TSP tag as increments of 4. However, technical staff noted the reliability and accuracy issue could be resolved if warranted by a longer term effort.
  • While the AVM and AVL technical assessments were not intended to quantify the accuracy of the demonstration functions, the limited assessments were valuable in providing staff with increased insights into the functions.

Automated Vehicle Monitoring:

In the limited "bug" test conducted by vehicle maintenance staff, the AVM data correctly reported the disabled 35 mph switch and radiator fan control switch, and observed values for oil temperature, oil pressure, cooling temperature, and voltage.

Automated Vehicle Location:

A small sample of demonstration onboard AVL data displayed on the Transit GIS map, indicated the location data were well aligned with the coach's scheduled route, and the path taken in off-route operation.

Issues identified in the analysis included:

  • The demonstration onboard AVL data identified a small percentage of bus stops which were located incorrectly on the GIS map.
  • Apparent problems in GPS signal reception on Third Avenue in downtown Seattle and in the tunnel, which did not impact the performance of the demonstration system in operation.
  • The on-/off-route status indicator was triggered immediately by significant changes in the direction of coach travel, but gradual changes in direction were slow to indicate a change in on-route status. In addition, off-route operation on a street parallel and in close proximity to the coach's expected route did not trigger off-route status. The Smart Bus system had many system parameters that could have been optimized to improve performance had the supplier been given an opportunity, as would be expected in a system wide implementation.
  • The demonstration system capitalized on frequent stops to improve the onboard AVL GPS processing. Additional evaluation and testing may be required for King County Metro service with few stops, such as express or night service.

Conclusion: Increased technical understanding for requirements development

The demonstration technical assessments provided staff with an opportunity to further their understanding of the technical attributes of the Smart Bus demonstration equipment, such as the onboard AVL and on-route status mechanisms, and their potential impacts on operations, as well as the APC and other demo equipment. With increased understanding, transit staff are better prepared to develop specifications and performance criteria for the future system procurement.

Conclusion: Technical issues and suggested next steps

The demonstration was helpful in identifying technical issues and suggesting future bodies of work including:

  • Explore testing requirements, process for system acceptance:

Issue: While several evaluation technical analyses were limited in scope due to time and staffing constraints, and the temporary nature of the demonstration, they were nevertheless, labor-intensive efforts.

Next steps: The assessments highlighted the need to explore methods for more rigorous AVL, AVM and other onboard systems acceptance testing, including establishing performance criteria, level of effort, and methods for processing the large volumes of data that would be required. The AVL testing approach should include a representative sample of transit service routes and locations, as well as potentially challenging areas such as hills, tall buildings and heavily wooded areas.

  • Onboard systems requirements development:

The GIS/AVL assessment has highlighted areas of emphasis for onboard systems requirements development to meet the needs and special conditions found in the King County Metro Transit environment including:

    • A downtown Seattle urban canyon;
    • The Downtown Seattle Transit Tunnel;
    • Off-route operation, such as short-term reroutes;
  • On-route status mechanism setting:

As adjustable features of the demonstration equipment, determining the optimum settings for the on-route status function will represent a challenging decision for system implementation. If the off route parameters were set too loosely, under some circumstances, a rerouted bus could be more likely to fail to recognize its off-route status, and continue to make stop announcements in error. If the parameters were set too tightly, the system, could be more likely to erroneously determine small changes in heading and distance traveled as off-route status, causing the system to cease automated stop announcements and displays.

Implications for adopting this type of technology include teaching operators what actions will trigger off-route status, and what coach conditions will require operators to make stop announcements.

  • GIS stop assessment:

Issue: The GIS/AVL assessment suggested areas of the Transit GIS map where more precision is required in mapping bus stops onto the GIS street network. The upcoming GIS transportation network upgrade is expected to improve the accuracy of the street network from approximately +/- 25 feet in incorporated King County and +/- 250 feet in unincorporated King County, to +/- 10 feet countywide. The increased accuracy is expected to contribute to more accurate placement of stops on the street network.

Next step: Additional analysis should be conducted to determine whether the GIS transportation network upgrade will resolve the issue of imprecisely mapped stops, or if a more precise method of locating bus stops on the street network will be required, such as determining location coordinates of bus stops in the field.

  • Urban canyon assessment:

Issue: The GPS urban canyon effect was observed in the analysis of onboard demonstration AVL data collected along Third Avenue in downtown Seattle.

Next step: As the urban canyon effect has been problematic in AVL implementations at other transit properties, an assessment to confirm the extent of an urban canyon in downtown Seattle should be conducted as soon as possible, and the results incorporated into the system requirements development effort.

Objective 2: Assess customer satisfaction with the automated announcements and interior signs.

Conclusion: Valuable feedback from customers for improved specifications

The demonstration project provided a valuable opportunity to obtain customer feedback on the proposed announcement and display functions for the OnBoard Systems Project requirements development effort. Most respondents to the transit rider survey and Sound Transit CAAC discussion participants reacted favorably to the demonstration functions and identified areas for further improvement.

Automated stop announcements:

  • The vast majority (89 percent) of survey respondents rated the interior next stop announcements as very or somewhat helpful, with over half of respondents (60 percent) rating the function as very helpful.
  • Overall, Sound Transits Citizen's Accessibility Advisory Committee and the Deaf-Blind Workgroup discussion participants appeared satisfied with the interior next stop announcements. Most participants rated the clarity, volume, and ability to hear the announcements over ambient noise, as good.

Issue: Timing of the stop announcements:

While most (87 percent) of survey respondents indicated the announcements of major stops and transfer points were made at the right time, 11 percent of respondents said the announcements came "too late," indicating an issue with the timing of the announcements. As an adjustable setting, the particular distance parameter setting used by the demonstration equipment for determining the timing of the next stop announcements will need to be examined in light of KC Metro's operating needs, and incorporated into OnBoard Systems project requirements development.

Automated next stop displays:

  • The vast majority of onboard survey respondents rated the next stop displays highly in helpfulness and visibility, and considered the information accurate. The next stop displays were the most preferred demonstration function, selected by 43 percent of respondents.

Issue: ADA customer needs

Participants in Sound Transit's CAAC discussion expressed concerns with the display lettering color, and the location and speed at which the information changed on the display. Issues with sign visibility were noted from glare and the placement of the sign over the front window of the 40-foot Gillig.

Automated exterior route and destination announcements

While generally rated as very or somewhat helpful by three-fourths of survey respondents, this function was considered less helpful than the interior next stop displays and announcements. Seventeen percent of survey respondents rated the announcements as "not at all clear." The function was least likely to be selected by respondents if they could choose only one feature, selected by 8 percent of respondents.

Issue: Gillig exterior speaker location

The mid-coach placement of the exterior speaker on the 40-foot Sound Transit Gillig coach is problematic for people with visual disabilities, as noted by CAAC discussion participants. The speaker location should be moved to near the front door to allow people with visual disabilities to orient themselves to the entrance of the coach and hear the announcement more clearly. Placing the speaker at a consistent location near the front door of all coaches would be more practical.

Conclusion: Continue to incorporate customer feedback into the design process

A continuing challenge for the project will be achieving the optimal balance in qualities of the announcements and displays: volume, display lettering color, brightness - to meet the varied needs of a diverse rider population. As the project continues to refine system requirements, customer feedback will continue to be critical in the design decision-making process.

Specifically, more work will be required to:

  • Refine the placement of the signs in the various fleet types to minimize glare and maximize visibility.
  • Find the optimum rate at which the information on the signs change, requiring more research. Ideally, the product eventually purchased will allow vehicle maintenance or other technicians to control the speed settings.
  • Find an alternative sign lettering format that meets the needs of people with disabilities, and is not distracting to other riders. Customer research and examination of designs used by other transit properties can help address this issue.

Objective 3: Obtain transit staff and transit operator feedback on the new functions: ease of use, advantages and disadvantages.

Conclusion: Mostly positive feedback from transit operator participants; unmet needs, but good potential for Vehicle Maintenance.

The demonstration project provided a valuable opportunity for transit staff to obtain hands on experience with the demonstration functions in their work environment. Most of the demonstration functions were well received by transit operators who participated in the demonstration debriefing. The generic, canned reports of the demonstration AVM functions were not useful to vehicle maintenance staff as provided, but showed good potential.

Transit operator feedback:

  • Automated stop announcements: Transit operators who participated in the demonstration debriefing reacted positively to this function, noting it allowed transit operators to focus on driving;
  • Automated exterior route destination signs changes: Transit operators who participated in the demonstration reacted very positively to this function, perceiving it as a timesaver for transit operators.
  • Automated public service announcements: Transit operators reacted positively to this function, describing it as useful, but difficult to access on the demonstration transit control head installed above the operator's compartment.
  • Exterior route and destination announcements: Operators had mixed reactions to the function. Some operators liked the announcements, as they saved operators the task of shouting the route number out the front door to waiting customers. Other operators found the announcements annoying and repetitive, as the announcement was repeated each time the front or back door opened in a bus zone.
  • Schedule adherence and on-route status: Transit operator debriefing participants had a negative reaction to these functions, perceiving them as not useful.

Vehicle Maintenance staff feedback on AVM functions:

  • Vehicle Maintenance staff found the data from the discrete input/outputs (I/Os) more valuable than the data from the J1708 network, such as oil pressure, which are generally available through gauges or a laptop.
  • The demonstration Mobile TA tools and AVM web reports did not include I/O data.
  • VM staff found the web reports cumbersome to use and difficult to read. The web report data were not available until the following day.

Objective 4: Identify issues, lessons learned, and recommendations for potential system wide implementation of the "smart bus" onboard equipment.

Over the course of the demonstration, participating staff identified a number of issues and suggested improvements if functions similar to those provided in the demonstration were implemented system wide. These observations also suggest future bodies of work for the On Board Systems project and related efforts. Among the more significant issues, lessons learned, recommendations and next steps identified in the course of the demonstration were:

Lesson learned: Gap between demonstration functions and VM needs

The demonstration offered an opportunity for Vehicle Maintenance staff to react to an actual AVM product prior to the development of user requirements. VM staff saw great potential in the functions, if they were more closely tailored to meet user needs to:

    • Collect information required by King County Metro VM staff, especially discrete input/output data;
    • Provide the ability to triage coaches needing repair as they return to the base;
    • Report detailed AVM data within an hour;
    • Generate reports in a user-friendly format.

Lesson Learned: AVM for coach repair triage

VM staff noted the primary value of implementing the Mobile TA Tools or similar function would be the ability to triage buses returning to the base according to their need for maintenance. The function would also need to indicate which component is out of tolerance, and include information from discrete I/Os. VM staff also had suggestions for the location of a Mobile TA tools or similar type of installation at the base.

Next Step: A large body of work highlighted by the demonstration will be an examination of the current vehicle maintenance coach lane assignment process and BO coach ("bad order" or coach needing repair) process to develop requirements for the Mobile TA Tools or similar function. Along with system requirements, a recommended revised business process will need to be developed.

Lesson Learned: Fewer keystrokes for operator controls

While the Smart Bus demonstration Transit Control Head (TCH) was not fully integrated with existing coach functions, such as the radio, the TCH provided a mechanism for testing operator reactions to potential new function control features. Transit operators participating in the demonstration found overriding the next stop announcements and destination signs, or operating the public service announcements required too many keystrokes. A key factor in achieving transit operator acceptance of a future system will be easy to use operator controls.

Lesson learned: Value-added features for transit operators

Transit operators participating in the demonstration debriefing perceived the automated stop announcements, the automated exterior route destination signs changes, and the automated public service announcements as demonstration functions that enhanced their ability to focus on safe coach operation and customer service. In contrast, the schedule adherence and on- route status functions were generally less valued by participants, and some found the repetitive nature of the exterior route destination announcements irritating.

Next step: Emphasize value added functions for transit operators in the design of the future system to increase the likelihood of the effort's operational success. Consider providing operators with some control over the exterior route destination announcements function, or other alternative for improving operator satisfaction with the function.

Lesson Learned: Continued operator involvement in DDU design

Although implemented for the short-term demonstration without KCM transit operator input, the demonstration Transit Control Head provided a physical product for transit operators to react to, and provide feedback for the development of requirements for the future Driver Display Unit. A key challenge for implementing the next generation of onboard technology functions will be the design of a user-friendly DDU, formally within the scope of the Smartcard project. To date, transit operator involvement has been incorporated in the initial DDU design efforts.

Next Step: Transit operator input throughout the DDU planning, design, and implementation efforts, will play a critical role in the operational success of the effort. Continue to involve transit operators in the DDU design development effort.

Issue: Stop sequence data quality

A systemic issue identified by staff who analyzed the demonstration data was a small percentage of errors found in the demonstration stop sequence data, originating from KCM sources. Due to discrepancies in the stop sequence data, technical staff involved in the demonstration had difficulty setting up the onboard stops database and using the data collected on the coaches. Although few in number, the stop data errors generated incorrect stop announcements and displays during operation of the demonstration coaches. The demonstration experience suggests that while the current data quality is high, onboard systems will require more stringent data quality as the data are used more widely throughout the agency, and any errors are highly visible to customers.

Next step: Providing the onboard systems with correct stop sequence data will be critical to successful implementation of the automated stop announcements and display, and APC functions. The Stop Information System (SIS) project, currently in progress, has the goal of addressing this issue, including determining a stop sequence data quality process, and staff roles and responsibilities.

Issue: APC process change

As noted in the APC assessment and technical staff debriefing, the implementation of smart bus integrated APC equipment will result in changes to the current APC process, notably elimination of some data post-processing, and increased volumes of data.

Next step: Along with integrated APC technical requirements development, a significant next step in the effort will be defining APC business process changes and staff roles in an onboard systems environment.

Issue: Data management process change

The demonstration has highlighted significant future efforts related to the King County Metro data management process changes required for onboard systems implementation including the:

  • Data collection and data management process for the stops and route database onboard the bus;
  • Implementation of the biweekly scheduling process and version control in an onboard systems environment;
  • Distribution of data from the onboard database to the KCM user databases, such as APC, fare collection;
  • Historical database design and maintenance issues such as database structure, data storage, data format, and data access.

Objective 5: Identify the potential long-term benefits of implementing smart bus technology.

Conclusion: Benefits to work environment and business processes identified through demonstration experiences.

Through their experiences with the demonstration project, King County Metro Transit staff were able to more clearly envision the benefits of the demonstration functions if implemented. Specific examples of the potential benefits of implementing smart bus onboard systems identified by staff during the demonstration project included:

Enhancement to transit operator work environment

Most transit operators participating in the demonstration debriefing perceived the automated next stop announcements, automated route destination sign changes, and automated public service announcements as enhancements to their work environment. According to the transit operators, the functions allowed them to focus on the operation of the vehicle and customer relations. In a work environment with a number of devices needing the operator's attention, the automated destination sign changes were "one less thing to worry about." A goal of the Onboard Systems and Smartcard projects is to streamline the operation of several devices used by transit operators: the radio, destination signs, and the proposed Smartcard and Security Camera systems, through an integrated Driver Display Unit (DDU).

More reliable APC equipment, more streamlined data processing

APC project staff anticipate that implementation of technology similar to the demonstration APC units would provide a number of benefits to APC data processing including:

  • Fewer repairs on APC units: The current mat based APC system, installed on the coach steps, is prone to damage from passenger foot traffic and water damage. An alternative technology such as the Smart Bus demo infrared light beam, is expected to be less prone to damage in the coach environment.
  • Easier repairs, less impact on data collection: New, modular APC equipment will allow technicians to more readily remove and replace the broken unit.
  • Less post-processing of data: A considerable level of effort is involved in processing APC data after it has been collected from coaches. APC staff anticipate a new APC system that integrates stop data with passenger counts will streamline the process.

Enhancement to customer's riding experience

While the Smart Bus announcements and displays were designed primarily to benefit riders with disabilities and to meet federal ADA requirements, other riders also found the functions useful. Approximately 90 percent of survey respondents found the interior next stop displays and the interior stop announcements helpful. While the demonstration transit routes did not constitute a representative sample of all King County Metro transit service routes and their riders, the survey results suggest a potential that other customers may perceive the functions as enhancements to their riding experience.

More efficient triaging of coaches needing repair

Vehicle Maintenance staff participating in the demonstration noted implementation of the Mobile TA Tools or similar function would allow them to identify coaches needing repair ("BO" coach) and park them were they can be easily accessed. Staff report BO coaches are frequently parked in a lane with other coaches, and accessing a BO coach requires that surrounding coaches are moved out of the way by VM staff. Participants estimate this happens at least once per shift at each base, taking approximately 30 minutes of staff time per incident.

Even if this scenario occurs only half as frequently as estimated by participants (and at six of the seven bases, as Bellevue Base has stall parking rather than lane parking) Vehicle Maintenance staff spend an estimated 31.5 hours per week or 1,638 hours annually accessing "buried" BO coaches. This represents a large potential time savings system wide if the process can be conducted more efficiently.

Accessible Vehicle Maintenance diagnostic data:

Through their experience with the demonstration AVM functions, participating Vehicle Maintenance staff envisioned the potential benefits of an AVM system if tailored more closely to staff needs. Quick access to diagnostic AVM data would enable staff to:

  • Manage fleet availability more effectively by identifying which coaches can go back into service and when;
  • Manage staff work flow more effectively by identifying the expected duration of a repair job and staff availability to perform the work;
  • Provide more preventive maintenance by utilizing the time saved from more efficient repairs.

Alternative for scheduling wheelchair lift preventive maintenance

Wheelchair lift preventive maintenance is currently scheduled by mileage. If an AVM function were implemented, VM staff suggested monitoring wheelchair lift use as an alternative method for scheduling PM. The potential benefit is possibly quite large if the reliability of the lifts were increased. 1,477 trouble calls were generated in 2001 as a result of wheelchair lift problems, requiring the dispatch of VM staff to assist a coach in the field.

Updated: Sept. 2002