Two of the systems measure spot speeds (the speed of vehicles passing a fixed point) and two measure travel times (based on identification of particular vehicles traveling the measured route). Travel times could be estimated based spot speeds but the accuracy of these estimates would depend on the variability of speeds between detectors.
In general, all the systems were found to record fairly accurate traffic characteristics. The systems were almost always accurate within 10 miles per hour for spot speeds or within 10 seconds for travel time. The exception to this was the travel time on Route 1-5 as measured by the Blufax system, though its errors could have been fixed testing for a biased error after the system is deployed and adjusting system to account for this error.
In order for a system to be portable it’s necessary that it be able to run on battery power. The batteries on all the systems tested required recharging at least once during the demonstration project. This problem could be addressed by using solar panels or more batteries. Some of the systems were moved by construction staff, as they were mistaken for traffic control devices. The LPR system had the most issues of reliability with several instances of the units shutting down, presumably due to overheating. Both of the travel time systems had challenges estimating travel times during periods of low traffic, as they only captured travel times for a portion of the passing vehicles (2 percent for LPR and 8 percent for Blufax).
The ease of installation and usability could be improved for some of the systems. The Blufax and iCone systems are simply set in place and switched on. The LPR system requires alignment of the cameras and calibration of the software. The Adaptir system requires alignment of the radar units and the RTMS unit.
Of the four systems, Adaptir is the only one integrated with communications and software to allow for real-time updates to the traffic management center. The Adaptir system also could have been configured to update the PCMS directly. The iCone system does have a communication component to allow for storage of the recorded traffic data on a central server, but it would take some work to provide real-time warnings to the traffic management center or to automatically update PCMS. The LPR and Blufax systems, as provided, were not integrated into a deployable system and were evaluated with post-processing of the data collected. However, these two systems could be integrated into a similar deployable configuration.
Further demonstration of portable ATIS systems should be done to better evaluate their effectiveness over a longer testing time period and possible a longer demonstration site (say 10-20 miles long). Testing at a site with no cellular phone service would be useful for verifying the usefulness in a rural setting. Future demonstrations should also be conducted using systems that have integrated control and communication capabilities that allow for automatic PCMS updating. This testing should be done at sites that have a high likelihood of experiencing delay so motorist response to the system can be evaluated.