Effectiveness of Different Approaches to Disseminating Traveler Information on Travel Time Reliability (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

40 Tasks 1 through 3 of this project examined previous literature, expert opinions, and state-of-the-art technology and future trends concerning the dissemination of travel time reliability information. In Task 4, building on the results of the first three tasks, the project team formulated a focused set of avenues of investigation to be conducted in Phase 2. During this task, the TTI team honed the research effort to ensure that a manage- able and realistic experimental design was developed for the remainder of the project. In reality, not enough time or dollars existed to examine every feasible combination of technology platforms, communication channels, message sets, and message formats to arrive at a single solution to the challenge of dissemi- nating travel time reliability information. Thus, Task 4 focused on summarizing the most important gaps that emerged from Tasks 1 through 3 related to the most effective means of communicating reliability information to travelers. Key research Issues The research team identified several important emerging research issues (which were addressed in Phase 2 of the proj- ect). The following items highlight those issues, the answers to which were pursued in the remainder of the project. Do Travelers Want Reliability Information? The central issue is to determine if there is market demand for reliability information. Transportation planners, policy- makers, and operations staff clearly find value in reliability data as a performance metric. Whether travelers find the same value is less obvious. Do Travelers Comprehend the Basic Idea of Reliability as Separate from Real-Time Travel Time Information? During Phase 1 of this project, the research team struggled to separate the two concepts—reliability information versus real-time travel time information—in discussions, writing, and planning. In discussions with the experts interviewed, Technical Coordinating Committee members, and even other Reliability program researchers, the team became aware that most people lapse into talking about real-time information when the topic is information dissemination. This is to be expected because most travelers have experience with free- way CMSs that display current conditions. The L14 team continued to correct people about this misunderstanding and stress the role of historical data in travel time reliability information. Do Travelers Comprehend the Variability Inherent in Reliability Information? The literature review showed, across several domains, how poorly people understand the concepts of variability, proba- bility distributions, and uncertainty. Do Travelers Comprehend the Average Travel Time? Some travelers may be making their trip time estimates on a near-worst-case basis, rather than on the average time the trip might take. If a regular commuter always travels in a narrow time window during peak hours, the overall average is not relevant for him or her. Thus, that commuter may have a different concept of typical or aver- age, or both. What Terms and Displays Can Best Communicate Average, Variability, and Reliability? The most easily comprehended system needs to use terms that are generated by laypeople, even if the terms are not tech- nically precise. C H a P T e r 5 Developed Avenues of Investigation

41 process that can require repeated queries to the system to exam- ine alternatives. That level of cognitive effort and interaction with a system is surely distracting. Here again, the system may need to make an opaque calculation that results in simple alter- native route advice. Since more states are banning all use of manual input devices—especially smartphones—the deploy- ment of these systems must be carefully considered. The Task 2 technology scan revealed that the future holds more automation and customization in smartphone and navi- gation system applications. Those tools may mitigate distrac- tion effects. Ultimately, en route use of in-vehicle systems was not tested in the human factor studies because the decision was made to replace the field operational test with the enhanced laboratory experiment. Do Travelers Want to Actively Seek Reliability Information (Pull) or Have It Sent Automatically (Push)? For both pre-trip planning systems and en route systems, some question remains whether reliability information should be pushed or pulled. Reliability information is calcu- lated on a historical basis and changes slowly over time, so the need for a push may not be urgent. Push data, instead, could be limited to real-time information. Users may want a quar- terly update or a push message when the reliability index of their selected route changes above some user-provided threshold value. How Can Multiple Data Sources and Factors Be Displayed for Reliability Information? Reliability algorithms use many factors to make travel time predictions, including weather, time of day, day of week, con- struction, transit system schedules, and so on. Do travelers need to know all of the factors, and do the factors need to be displayed at all? Even if all of the calculations are hidden or only available on secondary informational screens, travelers may develop trust in a system. Will Travelers Change Their Habitual Time of Departure and Route on the Basis of a Travel Time Reliability Information System? Humans are creatures of habit. In many ways, regular com- muters have done all of the reliability calculations in their head using their N = 1 data set. They know that when it rains, their trip usually takes five extra minutes; and if they leave at 7:40 a.m. instead of 7:30 a.m., the traffic is worse. What is not clear is whether travelers will develop enough trust in a sys- tem to make them change those regular habits. As the research project moved forward, the project team attempted to answer these questions within the framework of Phase 2, as illustrated in Table 5.1. Can Travelers Assign a Dollar Value to Reliability Information, or Incremental Changes in Reliability? Researchers need to understand whether travelers might be willing to pay a premium for reliability information. Under- standing the value travelers place on reliable trips could pro- vide insight into how agencies might prioritize design and operational changes to improve reliability on critical routes. Does Reliability Information Have a Role in Real-Time Systems? Algorithm designers can certainly use historical travel time information along with real-time data to improve the accu- racy of real-time predictions of travel time for a particular trip. From a traveler point of view, however, the role for his- torical data is unclear. If the real-time system says a trip takes X minutes, travelers may not care if it typically takes X to 10 minutes. Real-time data seem to trump historical data, as real-time data are most pertinent to today’s trip. Systems may need to be opaque about their use of historical data. Do Travelers Want or Need Reliability Information En Route, or Is It Mainly a Pre-Trip Planning Tool? Assuming travelers can comprehend reliability concepts and a lexicon can be developed, the question remains whether they would use the reliability information en route. The more apparent use is as a pre-trip planning tool to select recurring times of departure, routes, and modes. Thus, travelers may only access a system once when initially planning a commut- ing route or a particular trip to a new destination. Does a Traveler’s Desire for Reliability Information Change as a Function of Trip Purpose, Route, and Constraint? Trips are always made in a context. For constrained trips (e.g., the traveler needs to get to the airport), reliability information may be valuable and desirable for route choice or time of departure selection. However, for unconstrained trips (e.g., grocery shopping), reliability information may not be desired because the traveler incurs no penalty for being late or early. Can Travelers Safely Use an En Route, In-Vehicle System? Given the emerging evidence of driver distraction caused by in-vehicle devices—whether nomadic, such as cell phones and portable navigation systems, or native to the vehicle—traveler safety is an important question. Travelers may use reliability information en route to make route or mode shift decisions. Such decisions may involve a complicated decision-making

42 Table 5.1. Key Research Issues and Tasks Question Tasks 6-10 A1: Focus Groups Tasks 6-10 A3: Computer Survey Tasks 6-10 A3: Travel Behavior Laboratory Experiment Tasks 6-10 A3: Open-Ended Survey Tasks 6-10 B: Enhanced Laboratory Experiment Do travelers want reliability information? √ √ √ na √ Do travelers comprehend the basic idea of reliability, as separate from real-time travel time information? √ √ √ √ √ Do travelers comprehend the variability inherent in reliability information? √ √ √ √ √ Do travelers comprehend the term average travel time? √ √ √ √ √ What terms and displays can best communicate average, variability, and reliability? na √ na √ √ Can travelers assign a dollar value to reliability information, or incremental changes in reliability? √ na √ na √ Does reliability information have a role in real-time systems? √ na √ na √ Do travelers want or need reliability information en route, or is it mainly a pre-trip planning tool? √ na na na na Does a traveler’s desire for reliability information change as a function of trip purpose, route, and constraint? √ na √ na √ Can travelers safely use an en route, in-vehicle system? na na na na na Do travelers want to actively seek reliability informa- tion (pull) or have it sent automatically (push)? √ na na na na How can multiple data sources and factors be displayed for reliability information? na na √ na √ Will travelers change their habitual time of departure and route on the basis of a travel time reliability information system? √ na √ na √ Note: na = not recommended for this message; √ = suitable for this message.