This study used the HOWLATE (Heuristic On-Line Web-Linked Arrival Time Estimator) methodology to quantify the benefits of a prospective personalized pre-trip notification service in the Washington DC region. Based on 15 months (March 2000 through May 2001) of archived real-time travel time data collected from SmarTraveler.com, the study recreated 12 months of representative roadway travel outcomes on a model transportation network of 150 links and 55 nodes. The travel objective was to arrive on-time; therefore, to measure the difference between drivers using the pre-trip notification service and drivers not using it, yoked (paired) driving trials were recreated. For each simulated driving trial, one driver used the departure time and route recommended by the service, while the other driver did not. The ATIS non-user represented a commuter who maintained the same departure time and route choice day-after-day, while the ATIS user represented a commuter who used the pre-trip notification service to arrive on time. Based on the travel times and on-time arrival outcomes for 25.9 million paired driving trials, the travel time and on-time reliability of ATIS users and ATIS non-users were compared.
The simulation results were dependent on the estimated accuracy of archived real-time travel time data used to establish the baseline for ATIS non-user travel, and the ATIS non-user tolerance for late arrivals.
Overall, regular users of the pre-trip Advanced Traveler Information Service (ATIS) benefited significantly by reducing the frequency of early and late arrivals by 56 percent and 52 percent respectively. ATIS users were early 12 percent and late 2 percent of all trips, and ATIS non-users were early 28 percent and late 4 percent of all trips. When arriving early or late, ATIS users were on average 11.5 minutes early and 3.2 minutes late. ATIS non-users were 14.8 minutes early and 3.4 minutes late. These values constituted 22 percent and 7 percent reductions in the magnitude of early and late arrivals. ATIS commuters accomplished these savings by making different departure and/or route changes in nearly 60 percent of the simulated trips. Departure time changes were 13 times more frequent than route changes.
Approximately 20 percent of trips were made during the PM peak. ATIS proved marginally worse in reducing the magnitude and frequency of late arrivals and did not reduce in-vehicle travel time for these trips. During the PM peak, the average magnitude of delay was 3.4 minutes for ATIS users and 3.3 minutes for ATIS non-users. On average during PM peaks, ATIS users were late 4.2 percent of trips, and ATIS non-users were late 3.3 percent of trips.
Approximately 22 percent of trips were made during the AM peak and ATIS users had a 47 percent and 29 percent reduction in early and late arrivals respectively compared to ATIS non-users.
This study used field data collected from instrumented probe vehicles to evaluate the accuracy of the SmarTraveler data input into the model. The error assumptions associated with SmarTraveler data were somewhat optimistic; therefore, the overall benefit estimated may be somewhat higher than would be realized by using SmarTraveler.
See Also: On-Time Reliability Impacts of Advanced Traveler Information Services (ATIS), January 2001.
(Our website has many links to other organizations. While we offer these electronic linkages for your convenience in accessing transportation-related information, please be aware that when you exit our website, the privacy and accessibility policies stated on our website may not be the same as that on other websites.)