TFGAP Static Service.
Version of 15 October 2005.
Dave Barber.
TFGAP-A
1. Introduction.

Many intersections, especially those that are busy or dangerous, use traffic signals to inform drivers of when they may proceed, and of whether they must yield. These signals typically use some combination of dots and arrows in the colors red, yellow and green. Some lights flash (at about one cycle per second) when activated; others are steady.

Dynamic versus Static. At most intersections, the signals operate on a sequence of green, yellow and red lights -- under TFGAP known as dynamic service. Usually, the time to complete the whole cycle lies between 30 and 180 seconds, although with modern traffic-sensitive controllers the duration at any one intersection can fluctuate substantially.

For other intersections suffices a simpler plan, in TFGAP termed static service. This has two varieties: all-way static service presents flashing red lights to all drivers on all approaches, while some-way static service displays flashing yellow lights to some motorists and flashing red to others. Within either variety is occasionally included a steady green arrow. The word static is used because this display might remain the same for hours or years.

Some intersections are equipped only for static service, others are programmed to switch between static and dynamic according to the time of day or level of traffic, and yet others drop into static service only when some aspect of the control circuitry for dynamic service has failed.

Perennial versus Ephemeral. It is further possible, at one intersection, to operate in all-way service at certain hours with some-way at other hours. When the stopping or yielding rules for all approaches never change, the static service is termed perennial; falling into this category are virtually all intersections that are regulated solely by signs and pavement markings. On the other hand, when the rules do change from time to time, the static service is called ephemeral. This second category, practically speaking, requires signals: if requirements for stopping or yielding depend on time of day or other factors, explanatory signs would become too complicated for motorists to comprehend while busy driving.

One situation where ephemeral service is in order is near an elementary school, where an intersection could be an all-way stop at the beginning and ending of the school day when children are crossing the street, and a some-way stop at other times. Another example of ephemeral service displays one combination of signals at rush hour, and other signals during times of lighter traffic. This might operate on a fixed schedule, or be governed by a detection system that responds to the number of vehicles on the road. If the controller frequently changes the signals, the line between static and dynamic service begins to blur.


This report discusses static service in detail, along with changes to and refinements of traditional practice. Particularly, TFGAP offers substantial support for ephemeral service. Certain three-letter abbreviations for sections of traffic signals, decoded as follows, will be appearing often:

For example, FRD means flashing red dot and SGA means steady green arrow.

Most aspects of traffic light operation are discussed in MUTCD chapter 4D, with static service detailed in sections 4D.11 and 4D.12.


2. Placeholder Signals.

A placeholder signal is a traffic light, at an intersection, that is displayed to approaches during those moments in the signaling cycle when the drivers on those approaches have the same freedom and responsibilities as if there were no traffic light at all. One example of a freedom is the privilege of going straight without having to yield; one example of a responsibility is yielding to oncoming traffic when making a left turn; two examples of the signal are the SGD in dynamic service and the FYD in static. The placeholder needs to be lit whenever the other sections of the signal are unlit, with the result that no traffic light is ever completely dark unless it is malfunctioning.

The placeholder signal is not theoretically required for regulating traffic; a dark signal is technically sufficient. However, the placeholder offers practical advantages over the dark signal:

Placeholder signals, while beneficial, do not guarantee absolute safety. For instance, a traffic light controller could misbehave in such a fashion as to simultaneously show SGDs in all directions, resulting in collisions. Fortunately such malfunctions turn out to be rare, and problems that would otherwise lead to danger almost always do result in dark signals, at least on some approaches, so drivers get warning of the situation. For example, a burned-out light bulb assuredly results in darkness.

In dynamic service placeholder signals are important, with rarest exception, for all approaches to all intersections. By contrast, in static service placeholder signals may sometimes be omitted. For instance, suppose that at some intersection where a minor road crosses a major road, drivers on the major always get the FYD while those on the minor always see the FRD -- this is the classical "two-way stop". Removing the FYD shown to drivers on the major road is tolerable because the FYD light does not affect whether any driver stops, or how any driver yields to any other. From a different viewpoint, MUTCD 4K.5 reaches the same conclusion: at a some-way stop that is sign-controlled, FRD warning beacons can be mounted on any of the stop signs even if there are no FYD warning beacons on the major road.

The signals at railroad crossings have no placeholders, being dark except when a train is near. However, one measure of safety is provided by the battery backup required in MUTCD 8D.2. Of further, but limited, help are that any railroad crossing with a signal will also have the crossbuck sign, and that trains usually sound their horns when approaching a grade crossing. Although a green placeholder signal for railroad crossings has appeal, that green light may cause confusion when a grade crossing is very close to a signalized intersection, which has its own green lights.

Another lack of placeholder occurs in some parts of Europe, where signals are intentionally turned off at times of low traffic. These cases are resolved by signs posted at the affected intersections telling motorists what to do.


3. The Flashing Yellow Dot.

3A. The FYD versus the SGD. As mentioned above, the flashing yellow dot does not affect whether any driver stops, or how any driver yields to any other. The same can be said of the steady green dot. So exactly what is the difference? The conventional reply is that the FYD tells motorists to "use caution". Yet many motorists will use caution when approaching an SGD as well, because of the normal dangers of any intersection. Meanwhile, other motorists realize the the FYD imposes no legal requirements, so they pay it little heed.

A more specific response to the question is that the FYD at an intersection usually means that cross traffic is permitted to proceed after stopping and yielding. Meanwhile, the SGD says that cross traffic has a steady red signal that altogether prohibits going straight or turning left. Yet there is always risk of drivers running red lights, and with either the FYD or the SGD there is the possiblity that drivers from the cross road will be turning right on red lawfully. As a result, the distinction between the FYD and SGD is blurred.

Dealing with emergency signals for fire stations and the like, MUTCD 4F.3 acknowledges outright the equivalence of the FYD and SGD, reading "...The signal indication, between emergency-vehicle actuations, shall be either steady green or flashing yellow...". Further, MUTCD 4D.12 recognizes that a steady green signal can be compatible with static service: at an otherwise all-flashing intersection the MUTCD permits an SGA, if it is the sole component of a single-section signal.

In our view, two different signals are being used for essentially the same purpose. This introduces unnecessary complication, and TFGAP recommends that the SGD be used instead of the FYD in static service. To illustrate, the classical two-way stop would display the SGD to the major road, and the FRD to the minor. As a practical matter, replacing flashing yellow dots at intersections with steady green dots will require a number of years. After an additional period of perhaps a decade, motorists will become unaccustomed to seeing the FYD. Only then, with appropriate driver education, can the FYD be re-introduced in a new role discussed in the next section.

3B. The FYD and the FGD. Absent from traditional signaling is a light that tells drivers to yield. As the flashing green dot is otherwise unused, TFGAP recommends it for this purpose. Analogous to the FRD's equivalence to the stop sign will be the FGD's equivalence to the YIELD sign. A refinement is that the dot in either case can be replaced by an arrow. For instance, motorists on one approach to an intersection might be presented with both an FGA pointing to the right and an FRD. In this case, drivers turning right must yield, while drivers going straight or left must first stop and then yield.

A principle of TFGAP is that the clearance interval for a green light is always indicated by a yellow light. Additionally, the yellow should match the green; for example, a flashing green is cleared by a flashing yellow, a green arrow is cleared by a yellow arrow, et cetera. So TFGAP says that an FGD ought to be cleared by an FYD, but this raises a conflict with traditional usage. Because the FGD implies yielding, so should the FYD; but under traditional rules, the FYD does not tell drivers (at least those going straight) to yield. This difference of message may lead to crashes.

As explained in the previous section, the FYD will not be available for many years, hence traffic light phasing cycles will require a workaround for terminating the FGD. One plan is described in section 4B. After the workaround and education period, engineers can begin to install the FYD for clearance of the FGD.

The reader is cautioned that the SYD is not suitable for terminating the FGD, because the FGD requires straight-through drivers to yield, and the SYD does not.

3C. The FYD and the FBD. What about the FYDs currently used not at intersections, but as attachments to warning signs? They should be changed to flashing blue dots, as explained in TFGAP-B. This will ultimately leave the FYD to mean nothing but FGD clearance.


4. Examples of Configurations.

The examples normally call for identical dot signals for the opposite approaches of an intersection; the key reason is that many motorists assume that dots are so programmed. We do not know of a simple way to inform drivers when the signals are configured otherwise. Another matter is this: if traffic management at the intersection needs to be so sophisticated as to require different signals for opposite approaches, then the intersection may well need dynamic service rather than static.

The examples, because they involve sizable tables, open in separate pages for the reader's convenience. Within those tables, signals of multiple sections are shown horizontally, rather than vertically, in order to better fit the information onto a computer screen. By way of introduction, here are some typical signals:

Flashing
Steady

Many examples rely upon this simple arrangement of streets:

4A. Perennial Service.

4B. Terminating the FGD.

4C. Rotating the two-way yield or stop.


5. Alivial Information.

5A. General. In static service, motorists often benefit from knowing what traffic controls drivers on other approaches have -- it reduces waiting. We call such information alivial (from Latin ali-, other; and via, way). In a two-driver example, the first is waiting at a stop sign, and sees on the crossroad the second driver approaching her own stop sign. The first can proceed if he believes that the second is indeed planning to stop. On the other hand, if the first sees that the second has no stop sign, the first will likely wait until the second has passed.

In dynamic service, alivial information is of much less importance. Suppose a motorist is facing a signal more restrictive than an steady green dot, and is unsure of what other drivers will do. Under almost any phasing plan, he need merely wait a moment and then will receive a steady green signal.

The importance of alivial information is a probable reason that:

STOP and YIELD are the only signs meant to be observed from both front and back. The two sides are of course quite different in meaning: one side of each sign is regulative, while the other is merely informative.

5B. A Problem. When static service is not perennial but ephemeral, signs are difficult to implement. A case in point is an intersection that operates in dynamic service during the day, and switches to some-way stop at night. Such an junction cannot have ordinary stop signs posted, because they do not apply during dynamic service. A changeable stop sign might be considered, but the overall shape of the sign raises a problem. If the sign is octagonal, then even when the sign is turned off drivers on other approaches will see the back of the octagon and conclude that a stop sign is in force. If the sign is not octagonal, drivers on other approaches will not know that it is a stop sign even when it is activated. And a sign whose shape can be changed is likely too complicated to be feasible.

5C. A Partial Solution. A flashing red numeral, next to the flashing red dot, can fill in for the number-of-ways plaque. The letter A substitutes for the numeral at the all-way stop. In a convenient abuse of terminology, we sometimes refer to this letter as a numeral.

In static service, the signal might look like the following. Because one of the red sections is the plain dot, drivers who are unacquainted with this signal should at least figure out that they need to stop. According to MUTCD 4K.5, two red sections when arranged horizonally must flash together (except on railroad signals), and when arranged vertically must flash alternately.

   

When this signal is instead in dynamic service, it presents a familiar appearance:

   

If only one red dot is lit, the asymmetric display may cause motorists to speculate that the signal is broken. MUTCD 4D.16 permits duplication of signal sections, and red is the color most often duplicated.

Perhaps after some years, when motorists are widely familiar with flashing red numerals, the dot can be omitted:

   

5D. Limitations of the Partial Solution. The numeral is unable to communicate unusual arrangements of stop indications. It can be recommended without reservation in only the following situations, which fortunately happen to be the most common:

The term "four-way stop" is often imprecisely interpreted to mean "all-way stop". (This carelessness even appears in the option clause of MUTCD 2B.4.) Such error can lead to danger, as at a six-leg intersection with one major through road and two minor cross roads. If the major road has SGDs in both directions, and all approaches of the minor roads have flashing red, the intersection will be a genuine four-way stop. However, some motorists would misinterpret the numeral 4 as meaning an all-way, hence six-way, stop. Crashes may ensue. Safer is to use plain FRDs on the minor approaches, which effectively require drivers to study the intersection before proceeding.

The numeral is considered a partial solution because it cannot convey the subtleties of an intersection where opposite approaches have different signals. Case in point: at the intersection illustrated below, drivers approaching from the south, east and north must stop. By contrast, drivers approaching from the west are allowed to proceed uncontrolled because of a north-south railroad line only a few meters west of the intersection. The purpose of this three-way stop is to permit cars on the tracks to proceed without delay in case a train is nearing. A debatable issue is whether an FR3 displayed to motorists on the stopping approaches would be better than a plain FRD, because drivers facing an FR3 have no way of telling which of the other directions is uncontrolled.

Certainly one can devise a traffic light that informs a motorist of not only the signal for his own approach, but also of the signals for all other approaches. However, we do not know any way to do this without greatly increasing complexity, and consequently the risk of driver misinterpretation. The broader response is this: if efficiency or safety at an intersection in static service has become critical, then probably the best remedy is to convert the signals to dynamic service.