Meenakshy Vasudevan

MAKING THE MOST OF LIMITED DATA IN EVALUATING ADVANCED TRAVELER INFORMATION SYSTEMS BY EXPERIMENTAL RESAMPLING

Because of high data collection costs, analysts are commonly faced with the problem of limited data in the evaluation of intelligent transportation systems. How reliable are conclusions based on small samples? If limited data are available, how does one maximize their value? These questions were addressed to evaluate the potential benefits of prospective notification-based traveler information services used to deliver pre-trip travel time information to simulated drivers in a Cincinnati, Ohio, case study. In Cincinnati, travel time data were initially available for only 30 weekdays. An analysis that used this small data set indicated that an advanced traveler information system (ATIS) user would reduce disutility by 32% versus a comparable nonuser. However, since trip experiences on 30 weekdays may not characterize the typical experience of a commuter, conclusions drawn from the small sample may not accurately represent a more generalized assessment of the benefits of ATIS. Hence, an analogue of statistical resampling (experimental resampling) was applied to generate a large sample of days over which the effectiveness of ATIS could be evaluated. With experimental resampling, the reduction in disutility for an ATIS user was only 24%. It was concluded that experimental resampling provided a more reliable estimate of the benefit. To validate the claim, a more extensive study used 154 weekdays spanning a year. The validation analysis found that when compared with the small sample of 30 weekdays, the resampled cases were better predictors of the benefits for the large sample of 154 weekdays.

Comparison of Mobility Impacts on Urban Commuting: Broadcast Advisories Versus Advanced Traveler Information Services

This paper explores the effectiveness of relying on commercial radio as a source of traveler information and presents an approach to quantify mobility benefits from radio traffic advisories. The study, conducted for the Washington, D.C., metropolitan area, used an analytical technique called the heuristic online web-linked arrival time estimator to examine whether broadcast traffic advisories could have mobility benefits similar to a prospective notification-based traveler information service offering personalized estimates of travel times. Traffic reports were recorded from a local radio station and manually coded to translate them into a suitable format for analysis. Results from the analysis of 37 weekdays consisting of 4,410 advisories indicated that radio traffic advisories were less effective in improving traveler on-time reliability or reducing travel disutility than a service offering route-specific travel time reports. The simulated commuter receiving regular quantitative estimates of travel times on relevant roadways typically made more effective route and trip-timing decisions than the simulated commuter receiving comparatively incomplete, irregular, and vague advisories on prevailing congestion conditions from broadcast traffic reports. In contrast, the simulated commuter listening to radio advisories recorded similar benefits to those of the simulated control subject, who ignored all forms of traveler information. During the morning peak period, the simulated radio listener fared worse than the simulated control subject and recorded lower on-time reliability performance and higher travel disutility. During the afternoon peak period, the simulated radio listener had lower travel disutility but also experienced a nominal reduction in on-time reliability performance.

Quantifying Reductions in Commute Disutility from Traveler Information Services

An approach is presented for quantifying commute disutility measures. The approach was demonstrated through a case study conducted for the Washington, D.C., metropolitan area by using an analytical technique called the heuristic on-line web-linked arrival time estimation. “Commute disutility” was defined as having pretrip, en route, and posttrip components on the assumption that there is disutility associated with a commuters expectation of the trip before trip start, the en route trip experience, and the actual outcome of the trip. Three types of regular commuters were modeled: the nonuser who does not use any traveler information and two kinds of traveler information user–the radio listener who listens to commercial broadcast traffic advisories and the advanced traveler information service user who uses a notification-based service that provides route-specific travel time estimates. Analysis showed that compared to nonusers, traveler information users had lower commute disutility. They had fewer late arrivals at their destinations, and in fewer instances their trip expectation before trip start did not match their actual trip experience. They had fewer instances of feeling at intermediate waypoints along a trip that they were running behind schedule. They modified their trip start times or took alternate routes on more than 65% of the trips. This may have resulted in some disutility because of changes to the regular commute behavior, but they are more informed and therefore more confident of the potential trip outcome than is a nonuser.

Design and Development of Integrated Arterial Signal Control Model

This paper presents an arterial signal control system that provides arterial progression while optimizing signal timing plans at each intersection along the arterial. The proposed system is divided into two levels: progression control level and intersection control level. At the progression control level, arterial progression is provided by maximizing bandwidths using a modified version of the multiband model. In this study, unlike most bandwidth maximization models, queue clearance time and minimum green time are not prespecified, but are computed as a function of the existing queue at an intersection. At the intersection control level, signal timings are optimized for each intersection by minimizing a weighted combination of vehicle queue lengths, intersection control delays, and vehicle stop times, subject to bandwidths generated at the progression control level. The effectiveness of the system was evaluated through a case study conducted with CORSIM. Results showed that the proposed model was superior...

Predicting Congestion States from Basic Safety Messages by Using Big-Data Graph Analytics

In a connected-vehicle environment, wireless subsecond data exchange connects vehicles, the infrastructure, and travelers’ mobile devices. These data have the promise to transform the geographic scope, precision, and latency of transportation system control; fulfillment of that promise could result in significant safety, mobility, and environmental benefits. However, the new data influx also has the potential to overburden legacy computational and communication systems. Although connected-vehicle technology can facilitate ubiquitous system coverage, the existing prediction methods, computational platforms, and data management methods are insufficient to process the data within a reasonable time frame for real-time predictions. An investigation of the ways in which advanced (big-data) analytics might be applied to realize the full potential of connected-vehicle technology is particularly relevant now as this technology evolves from research to deployment. This paper presents an approach combining big-data graph analytics with high-performance computing to predict traffic congestion by analyzing nearly 4 billion basic safety messages generated by the safety pilot model deployment conducted in 2012–2013. This paper provides an alternative approach for predicting congestion in 30.5-m segments anywhere on the network at 1-min intervals 30 to 60 min before actual congestion over a time window of 1 h. Despite sparseness of data, the proposed framework predicted highly congested locations 40% of the time. Severity of congestion was predicted with an accuracy of 77%. This combination of rapid computation and predictive accuracy may provide significant value in future real-time decision support systems that leverage connected-vehicle data.

Mobility and Commute Disutility Effects of 511 Deployment in Salt Lake City, Utah

The mobility and disutility-reduction benefits to users of the 511 advisory service currently deployed in Salt Lake City, Utah, were evaluated through a modeling study conducted using the HOWLATE (Heuristic On-Line Web-Linked Arrival Time Estimation) methodology. The benefits of the existing 511 advisory service to seasoned users of the service and the potential additional benefit to providing travel-time estimates via the 511 service were evaluated. Data were provided by the Utah Department of Transportation. Four types of commuters were modeled: (a) a nonuser, who ignored traveler information; (b) a 511 advisory user, who made use of the 511 advisory service currently available to commuters in Salt Lake City; (c) a 511 travel-time user, who made use of a hypothetical 511 service that provided travel-time estimates; and (d) a dynamic message sign (DMS) user, who made use of DMS deployed along 20 mi of I-15. Analysis showed that the existing 511 advisory service resulted in mobility and disutility-reduction benefits to seasoned commuters. Compared with the nonuser, the 511 advisory user was able to reduce peak-period late arrivals by 14%. By supplementing the existing 511 advisory service with DMS, late arrivals for current 511 advisory users can be further reduced by more than 10%. Providing travel times for the entire network over the 511 service can reduce peak-period late arrivals for current 511 advisory users by nearly 30%.

A neural network approach to constructing commercial broadcast traffic advisories

This work presents a method for constructing an archive of broadcast radio traffic report content from Web advisories using neural networks. Broadcast traffic reports are free and widely used as a source of traveler information. However, there has been no study done to establish what impacts, if any, these traffic reports have in terms of improving listener travel reliability. We developed an analytical technique to quantify travel reliability impacts and conducted a preliminary case study for the Washington, DC, metropolitan area, using radio traffic reports recorded from a local radio station and manually coded for 37 weekdays. However, as coding of radio traffic reports is highly labor-intensive, we used neural networks to construct a database of radio traffic advisories from an existing archive of Web traffic advisories. This paper presents the model developed using feed-forward neural network with back propagation of error that can, given a list of Web advisories, predict roadway segments that would also have an advisory mentioned on the radio. The overall accuracy during the morning peak period was 72%, implying that a commuter listening to constructed advisories would have a 72% chance of listening to an actual advisory mentioned on the radio. During the afternoon peak period, the accuracy was 78%. The missed prediction rates in the morning and afternoon peak periods were 28% and 23%, respectively. Given that we can construct a full year of radio traffic advisories we are able to conduct a more representative study for a longer period of time since traffic conditions on 37 weekdays cannot be used to generalize typical trip experiences of a commuter. Thus, neural networks proved to be a viable low-cost approach to solve the problem of lack of data.