Mode of Flashing for Malfunctioning Traffic Signals


Peter Jenior, Graduate Student, Georgia Tech
Justin Bansen, Graduate Student, Georgia Tech
Valerio Oricchio, Graduate Student, Georgia Tech
Dr. Michael Hunter, Associate Professor, Georgia Tech
Dr. Michael Rodgers, Principal Research Scientist, Georgia Tech

Project Overview:

At a signalized intersection operating in flash mode, signal indications may flash red on all approaches (red/red flash) or flash yellow on the major road and flash red on all other movements (yellow/red flash).  While the meaning of yellow flash and red flash is regulated by the legal code of each state, drivers are generally required to stop at a flashing red signal indication and proceed with caution at a flashing yellow indication.  Among the possible reasons for entering flash mode, malfunction flash is unique in that it is a failsafe mode of operation that cannot be entirely eliminated.  The primary focus of this study is to propose a policy for the selection of malfunction flash mode at an intersection. 

Agency Survey

Two surveys of officials responsible for traffic signals were conducted – one of all Georgia agencies and one of selected agencies nationwide.  While yellow/red flash was the dominant flash mode, several agencies did report the use of only red/red flash.  Overall power interruption, lightning, and equipment malfunction are the three major causes of malfunction flash reported.  However, the percentage of flash occurrences attributable to any one of these varies widely from one jurisdiction to another.  It was also seen that most agencies rely heavily on citizens and law enforcement officials to identify signals that have gone into flash. 

Operations in Malfunction Flash

Data collection began in May 2005 and ended in December 2006.  A total of 43 instances of malfunction flashing operation in the Atlanta region were captured: 34 intersections under yellow/red flash and 9 intersection under red/red flash.  Due to the travel patterns of those involved with the study, most of the intersections were in highly urbanized areas.  For each vehicle through the intersections, the lane, stop time (if the vehicle stopped), departure time, and movement were recorded.  Data for 55,788 vehicles were recorded in this study. 

During field observations it was noted that a considerable percentage of drivers choose to stop when facing a flashing yellow indication, or choose to not stop at a flashing red indication.  The percentage of through vehicles stopping at an intersection when facing a flashing yellow indication ranged from 0 to 76 percent.  This is a major departure from previous operation studies, which assumed that no vehicles would stop when facing a flashing yellow indication.  Under low side street volume conditions this assumption may hold true; however, as the major street and minor street volumes increase, the intersection operation begins to more closely mirror that of an All-Way-Stop-Control intersection.  The minor street to major street volume ratio, minor street volume, presence of a minor street vehicle, and functional classifications of the roads at the intersection were all found to be correlated with percentage of major street vehicles stopping. 

The percentage of major street vehicles choosing to stop at an intersection operating in red/red malfunction flash control ranged from 78 to 93 percent.  This is also a departure from previous studies, which assumed that all drivers would stop at a flashing red indication.  No variable studied was a good predictor of this stopping rate.

Overall, both yellow/red and red/red flash provide reasonable performances at relatively low major and minor street volumes and both break down at higher volumes.   The operational advantages of yellow/red flash over red/red flash tend to be significant only where the demand and functional classification of the mainline dominate the cross street.


This study demonstrates that from the driver’s perspective malfunction flash mode can result in highly unpredictable traffic operations.  There is no “safe” malfunction flash mode as yellow/red flash often leads to confusing behavior among drivers and red/red flash has a relatively high instance of drivers failing to stop.  Clearly malfunction flash is undesirable and every effort should be made to reduce its occurrence and duration.  However, malfunction flash mode cannot be entirely eliminated, thus traffic engineers must choose a malfunction mode – yellow/red or red/red. Based on the results from this study it is concluded that RED/RED flash be the primary mode of malfunction flash

Yellow/Red Malfunction Flash Exceptions

It is recognized that if red/red becomes the standard mode of malfunction flash, there may exist limited scenarios in which yellow/red flash would still arguably be the preferred flash mode.  For example, at the intersection of a sufficiently large, high volume or high speed road and a sufficiently small, low volume or low speed road, few drivers choose to stop at a flashing yellow signal (e.g. a high speed rural highway).  As part of this study little data was collected at red/red flash controlled intersections where one road had a significantly higher volume, although the data that was collected leads to the concern that significant levels of major street failure to stop violations may occur under such circumstances.  Violations of the flashing red on the mainline could become a significant hazard.  Further study should be undertaken on operations during flash at these intersections.

Sponsor: Georgia Department of Transportation (RP 05-16)