HAWK Thread

[Dirt Roads]

A preliminary report from back when HAWKs were introduced in Phoenix indeed showed a reduction in the total number of rear-end crashes and basically no change in the total percentage.  Details are shown below:

Safety Effectiveness of the HAWK Pedestrian Crossing Treatment – July 2010

Rear-end is the most common crash type for all types of intersection control, representing between 40 and 60 percent of the crashes for a given type of traffic control at the intersection. After rear-end, the most common crash types were left-turn and angle.

The dataset shows rear-end as the most common manner of collision for the HAWK intersections in both the before and after periods (illustrated in figure 7 and figure 8). The distribution of IR crashes at the HAWK sites before the HAWK was installed included rear-end (55 percent), angle (13 percent), left-turn (15 percent), and pedestrian (5 percent) crashes. After the installation, the greatest changes in the distribution of crash type were an increase in angle crashes to 19 percent and a decrease in pedestrian crashes to 1 percent.

[...]

A preliminary review of crash type at the HAWK sites indicated a reduction in rear-end crashes, which is not typical when a higher level of control is implemented at an intersection. A potential reason for the reduction in rear-end crashes is that drivers behind the initial vehicle that has stopped for a crossing pedestrian can view the traffic control device without needing to see the pedestrian, who may be obscured by the lead vehicle. Additional research to investigate the changes in crash patterns at the HAWK sites should be considered.

https://www.fhwa.dot.gov/publications/research/safety/10042/10042.pdf

Please be careful highlighting the data.  It is likely that most of the incident/crash data related to "rear-end collisions" are unrelated to activation of the HAWK signal.  As with all safety data studies, the entirety of the data is used to calculate any trends.  But that doesn't mean that the "trend" is related to "causal changes" (in this case, implementation of a bunch of HAWK signals). 

The specific data itself is suspect.  IR means intersection-related data, whereas ISN means collision data flagged with the same intersecting street name (which should include data from that same intersection).  But I'm presuming that many of the HAWK applications were mid-block, meaning that there is an inexact correlation between the data and the HAWK applications.

What is quite important from this data is shown in the line above the highlighted one.  Overall pedestrian-related collisions from nearby locations (ISN) dropped from 27 (roughly 4% of the total) to 4 (roughly 1% of the total) after installation of over 60 HAWK installations in Tucson prior to the year 2006.  Table 4 indicates that is an 83% drop over the study period.  That data is influenced by the HAWK installations, but doesn't necessarily correspond to the HAWK installations themselves.  Note that there was also a drop in pedestrian-related collisions at signalized intersections.  Table 4 in the FHWA document indicates that there was an 18% drop in pedestrian-related collisions at signalized intersections (IR) over the same period.  (Some of that may have been impacted by HAWK installations as well, so don't go down a tangent there either).

Anyhow, the data is sufficient to demonstrate a significant decline in pedestrian collisions near HAWK installations.  But I doubt that anything else can be directly concluded from this study data.

[GaryV]

^ I don't recall where it was, but recently I went past a "New Signal Ahead" sign with red flags. Got to the signal, and they hadn't been turned on yet, maybe with black covers but I couldn't tell. Ugg. I know it's probably 2 different contractors, but still, you'd think they could coordinate.

[jakeroot]

I'm personally sold on the opposite problem: virtually no one stops at dark HAWK signals (because they shouldn't). This is good, of course, as they shouldn't. But is it potentially training drivers that not all dark signals require a stop? And when I say "signals", I mean things that are overhead or on the edge, and that light up with colors to indicate what to do...I assume that's how most drivers interpret it, anyways.

I don't like the dark signal rule anyway.

For example, if a new signal is going in, most agencies either turn the signals away from the road or put bags over the signals.  However, if I'm driving at 55 mph after dark, then it is not apparent if the signal is turned, bagged, or simply not functioning.  (And yes, this came up for me.  There's a specific signal I'm thinking of in my area that was recently installed along a 55mph stretch of road, and I couldn't tell until I was practically right under it.)

My preference would definitely be for fewer dark signals. But then, I don't know what rule we could implement in its place. Or what actions could be taken to ensure we never have dark signals; HAWKs and ramp meters certainly aren't helping the situation.

[kphoger]

But then, I don't know what rule we could implement in its place.

Agreed.

HAWKs and ramp meters certainly aren't helping the situation.

How many drivers are getting rear-ended at dark ramp meter signals because they thought it was a non-working stoplight?

[Scott5114]

Scott5114, the reason I bring up the issue of the public not understanding beacon vs. traffic control signal is because different driver conduct is expected when approaching a dark signal depending on the type.

Drivers are expected to treat a malfunctioning dark traffic signal like a stop sign. But they are permitted to drive thru a dark HAWK. So if they don't know the difference between the two types, how will they know what to do when they approach a dark HAWK signal. They may stop thinking it's a defective traffic control signal and maybe get rear ended for trying to do the right thing.

I don't disagree, but I'm pointing out that that difference in operation is intentional to snake through a loophole in the rules.

Wouldn't it have been a lot easier to just add a line to the 2009 MUTCD that made HAWK-eligible crosswalks warrant a traffic signal? That's always what's bothered me about them.

Possibly...but I think there's some conservatism toward watering down the signal warrants, lest cities use signals for speed control. (They already have that problem with stop sign warrants.)

I don't like the dark signal rule anyway.

For example, if a new signal is going in, most agencies either turn the signals away from the road or put bags over the signals.  However, if I'm driving at 55 mph after dark, then it is not apparent if the signal is turned, bagged, or simply not functioning.  (And yes, this came up for me.  There's a specific signal I'm thinking of in my area that was recently installed along a 55mph stretch of road, and I couldn't tell until I was practically right under it.)

The trend toward reflective backplates makes dark signals more obvious in the dark. Although you still have no way of knowing if they're bagged or not (unless they put the bag over the backplate too, which I've never seen). At least it still shifts the balance more toward "fail safe" than "fail deadly".

[SignBridge]

Well Kphoger, I hope you and the FHWA are correct and that there won't be too much public confusion over this issue. And no, I'm not aware of any rear-end collisions in the circumstances I suggested, but then there are not yet any HAWK signals in my county that I know of. 

Re: dark ramp metering signals, in my area there are yellow flashing lights with a yellow diamond shaped sign ahead of the actual signals. Drivers know when the yellow lights are flashing the ramp signals are in operation.

[kphoger]

(unless they put the bag over the backplate too, which I've never seen)

I don't actually consider it to be a "fail safe", though, anyway.  If I'm driving on a 55mph road (in the specific case I'm thinking of, in fact, it's in farmland between towns) and I come across a dark signal, I'm probably just going to assume it's a new installation and isn't operational yet.  It would actually take a fair level of certainty otherwise to make me come to a stop.

Re: dark ramp metering signals, in my area there are yellow flashing lights with a yellow diamond shaped sign ahead of the actual signals.

Here's a yellow diamond shaped sign.  I suppose they could put a yellow flashing light on top, if that would make you happy.

[jakeroot]

The trend toward reflective backplates makes dark signals more obvious in the dark. Although you still have no way of knowing if they're bagged or not (unless they put the bag over the backplate too, which I've never seen). At least it still shifts the balance more toward "fail safe" than "fail deadly".

I would say bagging the entire signal, backplate and all, is more common in my area than not. Both BC and WA.

HAWKs and ramp meters certainly aren't helping the situation.

How many drivers are getting rear-ended at dark ramp meter signals because they thought it was a non-working stoplight?

Didn't I just say that I thought drivers not stopping at dark ramp meters and HAWKs was the bigger issue, not drivers accidentally stopping and getting rear-ended?

My bigger concern remains the entire concept of intentionally dark signals (or "beacons" as they prefer they're called), further cementing, for some, that dark signals of any kind can simply be ignored. Because, well, they can. But not all of them.

[Dirt Roads]

^^^^
I'm already seeing a generation that gets confused at dark traffic signals.  Instead of stopping, they seem to think that the "main road" has the inherent right-of-way.  We've got a signal location closeby where the "main road" headed to the Interstate crosses US-70.  During several power outages over the past few years, folks on both of these roads think they have the inherent right-of-way when the signal is dark.  I've even heard someone blast their horn at the car ahead who correctly stopped at the "four-way stop" when they didn't plow through the intersection.

[mrsman]

Oh wait, these aren't traffic control signals as defined in the Manual. They are Beacons which are a different set of rules.

Yeah right; like the average driver knows the technical difference......

This is further reason that they should only be used mid-block, but drivers encounter dark beacons away from intersections all the time.  Think of a stoplight warning sign with an attached double yellow beacon:  some of them remain dark until the light is about to change.

Kphoger, the difference is that the stoplight warning beacon is just an advisory/caution beacon. Whereas the HAWK is a regulatory signal just like a normal traffic light, so in my opinion the FHWA shouldn't be calling it a beacon. They define it as a beacon because it doesn't meet their definition of a traffic control signal which operates continuously alternating the right-of-way between at least two approaches.

And again if the Feds think the average driver is going to make that distinction between different types of traffic lights, well good friggin' luck. It'll be interesting to see how many drivers stop at dark HAWK signals at night thinking it's a defective traffic light that's completely out.

Nobody expects the general public to care about the beacon/signal thing. The HAWK was intentionally designed to fall under "beacon" rules because if it fell under "signal" rules it would be subject to signal warrant requirements. Those warrant requirements are, I believe, precisely why a lot of cities couldn't install standard R-Y-G ped signals before the HAWK was invented (around here, the only places I've seen them are immediately adjacent to schools).

I'm guessing either Los Angeles doesn't care about the signal warrants because they have the political clout to not particularly care what FHWA/NCUTCD wants, authority granted by the state of California or the California MUTCD to ignore the federal signal warrants, or enough traffic to more consistently meet the signal warrants than other cities do.

Obviously the real solution is to correct the signal warrants to make R-Y-G signals for pedestrians easier to legally install, but hell, we're, what, a year overdue as it is for MUTCD 11e now?

Wouldn't it have been a lot easier to just add a line to the 2009 MUTCD that made HAWK-eligible crosswalks warrant a traffic signal? That's always what's bothered me about them.

In a word, yes.  The MUTCD should be adjusted to allow for easier warrants for pedestrian crossing signals.  With that said, I have a few caveats.

The pedestrian crossing signal warrants should be easier than that at a regular intersection.  But the crossing traffic at such signals should only be for pedestrian use.  This means that the crossing should either be mid-block or just in front of a side street that is RIRO. [The side street can only turn right onto main street, and a right turn would not interfere with  peds crossing the main street.]  In my view, the warrants for this type of pedestrian-restricted signal should be those of a crosswalk.  If we can expect a pedestrian to cross at the location without the help of any traffic control device, then adding this type of device would only make such crossing safer.  The key is that the flashing red (Los Angeles) or flashing yellow (PELICON) feature would allow traffic to continue, in the same way that traffic would continue at an unsignalized crossing, once the pedestrians are out of the way.  A normal traffic signal must meet warrants, because we have to gague the number of vehicles on the side street that need to cross (or turn left) and thus need to stop the flow of traffic on the main street for a considerable period of time.  For a pedestrian signal, the amount of time that traffic is stopped would be equivalent to the amount of time that a driver would be doing at an unsignalized crossing, assuming that the driver is giving the pedestrian the appropriate, legal, right of way.  In effect, the pedestrian signal doesn't actually change the flow of the main street - it simply makes the pedestrian crossing safer, with the benefit of the red signal.

At the same time, once pedestrian signals are implemented along a corridor, there has to be standards as to their placement.  Too many of even these types of signals would degrade traffic.  The traffic engineer needs to use their judgment to determine the proper frequency for all types of signals.  In Manhattan, it seems that there is a signal every 250 feet - the cross streets are every 1/20 of a mile.  And there are signals in Midtown, even when the cross street isn't there.  (Example: Madison Ave has no cross streets between 112th and 115th, yet there are two mid block pedestrian signals in the place of 113th and 114th.  And like nearly all NYC signals, they are not actuated.  They will turn red even if there are no pedestrians present.  And these signals do not have a flashing red feature like HAWKs.  I cannot anticipate seeing any type of specialized pedestrian signal in NYC any time soon.).  For most other cities, perhaps 500 feet is a good standard, and for many suburban roads, I can see the limitation on frequency being anywhere betweeen 1000 and 3000 feet.  The key question is how far do we expect pedestrians to walk to a safe crosswalk and site the crosswalks, crossing signals, and bus stops accordingly.

And once the warrants are adjusted, these crossings should be treated as signals, not beacons.  Signals with different warrants, since their purpose is ped/bike protection, not vehicular movement.  And my strong preference is to avoid intersections (which basically would mean that some side street traffic would take advantage and side street traffic control should only occur with a fully warranted signal) and to use either the Los Angeles or Pelicon variants, which I believe are simply more intuitive and safer than HAWKs.  In all cases, the signals should only be activated when a pedestrian pushes the button, and should allow a flashing red or yellow after a few seconds of solid red, to account for faster pedestrians who get out of the way and walk more quickly than 3.5 ft/sec.  There is no need to hold up traffic, once the pedestrians are out of the way.

[kphoger]

Didn't I just say that I thought drivers not stopping at dark ramp meters and HAWKs was the bigger issue, not drivers accidentally stopping and getting rear-ended?

Yes.  Sorry, I didn't mean my reply as an argument against you.  I was merely pointing out that there are already signal-like installations that go dark where people aren't expected to stop–and yet rear-end crashes aren't happening there.

I'm already seeing a generation that gets confused at dark traffic signals.  Instead of stopping, they seem to think that the "main road" has the inherent right-of-way.  We've got a signal location closeby where the "main road" headed to the Interstate crosses US-70.  During several power outages over the past few years, folks on both of these roads think they have the inherent right-of-way when the signal is dark.  I've even heard someone blast their horn at the car ahead who correctly stopped at the "four-way stop" when they didn't plow through the intersection.

When do you think this generation began the behavior?  I've personally been seeing it for twenty years, which is as long as I've been driving in urban environments.

[jakeroot]

I'm already seeing a generation that gets confused at dark traffic signals.  Instead of stopping, they seem to think that the "main road" has the inherent right-of-way.  We've got a signal location closeby where the "main road" headed to the Interstate crosses US-70.  During several power outages over the past few years, folks on both of these roads think they have the inherent right-of-way when the signal is dark.  I've even heard someone blast their horn at the car ahead who correctly stopped at the "four-way stop" when they didn't plow through the intersection.

When do you think this generation began the behavior?  I've personally been seeing it for twenty years, which is as long as I've been driving in urban environments.

Kind of begs the question as to why it's not already the law. Isn't there like a boulevard rule that could take its place?

[kphoger]

It's kind of like how a lot of people don't yield to the right at uncontrolled intersections, instead assuming the "main road" has right of way.  In fact, I think a handful of states even have that as the law, but most don't.

[Dirt Roads]

I'm already seeing a generation that gets confused at dark traffic signals.  Instead of stopping, they seem to think that the "main road" has the inherent right-of-way.  We've got a signal location closeby where the "main road" headed to the Interstate crosses US-70.  During several power outages over the past few years, folks on both of these roads think they have the inherent right-of-way when the signal is dark.  I've even heard someone blast their horn at the car ahead who correctly stopped at the "four-way stop" when they didn't plow through the intersection.

When do you think this generation began the behavior?  I've personally been seeing it for twenty years, which is as long as I've been driving in urban environments.

That's about when I began seeing this here, as well.  Not all of those drivers are/were young.  There just seems to have been a shift in society that the motor vehicle rules apply to everyone else.  (Note that is also about the same time that my career got routed over from transportation safety into transportation safety regulation.  So I might have gotten old and crotchety at the same time).

[GaryV]

^ I don't recall where it was, but recently I went past a "New Signal Ahead" sign with red flags. Got to the signal, and they hadn't been turned on yet, maybe with black covers but I couldn't tell. Ugg. I know it's probably 2 different contractors, but still, you'd think they could coordinate.

I went thru this intersection again today. The traffic light was working now - flashing red in all 4 directions. It's on 14 Mile Road at the exit of a furniture store parking lot; on the side opposite the store there is some new facility but I didn't see the details.

I stopped for the flashing red; most drivers were not stopping.

[Scott5114]

I'm already seeing a generation that gets confused at dark traffic signals.  Instead of stopping, they seem to think that the "main road" has the inherent right-of-way.  We've got a signal location closeby where the "main road" headed to the Interstate crosses US-70.  During several power outages over the past few years, folks on both of these roads think they have the inherent right-of-way when the signal is dark.  I've even heard someone blast their horn at the car ahead who correctly stopped at the "four-way stop" when they didn't plow through the intersection.

When do you think this generation began the behavior?  I've personally been seeing it for twenty years, which is as long as I've been driving in urban environments.

Kind of begs the question as to why it's not already the law. Isn't there like a boulevard rule that could take its place?

The problem is that at a number of intersections, which road is the "main" road can be not at all obvious, and the answer may even differ from person to person. Case in point, this intersection in Springfield, Missouri. Which road is the main one? I would say Kimbrough is, but that's because I lived there for a year and used that road enough to get to know the traffic patterns. To someone from another city, it would appear entirely inscrutable. If we were to meet at that intersection and both think we're on the main road . . .

And then there's the problem of what to do when two main roads intersect. You might have a different rule for those instances, but again, it requires everyone passing through the intersection to know, without ever having traveled on the intersecting road, that it is a main road.

I think Europe solves this problem by setting it up so that functioning stoplights legally overrule all other traffic control devices at an interchange, so they can provide supplemental stop/yield/you have priority signage at each intersection. If the stoplight goes kaput, then control of the intersection "falls through" to the supplemental traffic control signage.

Short of that, though, "always stop at an intersection with darkened lights" is the safest option.

[SignBridge]

Well Scott5114, many HAWK signals are installed at or extremely close to intersections, contrary to MUTCD advice. What do you suggest drivers do in those situations?

[kphoger]

The problem is that at a number of intersections, which road is the "main" road can be not at all obvious, and the answer may even differ from person to person. Case in point, this intersection in Springfield, Missouri. Which road is the main one? I would say Kimbrough is, but that's because I lived there for a year and used that road enough to get to know the traffic patterns. To someone from another city, it would appear entirely inscrutable. If we were to meet at that intersection and both think we're on the main road . . .

Some state laws likewise get a bit wonky when it comes to uncontrolled intersections.

For example, in Maryland, uncontrolled intersections have the typical yield-to-the-right rule–but not if you're entering a paved road from an unpaved road, in which case traffic on the paved road has the right of way no matter which direction it's coming from.

[Scott5114]

Well Scott5114, many HAWK signals are installed at or extremely close to intersections, contrary to MUTCD advice. What do you suggest drivers do in those situations?

Petition their DOTs to follow the damn MUTCD.

Seriously, "intersection not compliant with the rules causes problems" is not an indictment of the existing rules–it's an illustration of why the existing rules should be followed!

[SignBridge]

Well Scott5114, many HAWK signals are installed at or extremely close to intersections, contrary to MUTCD advice. What do you suggest drivers do in those situations?

Petition their DOTs to follow the damn MUTCD.

Seriously, "intersection not compliant with the rules causes problems" is not an indictment of the existing rules–it's an illustration of why the existing rules should be followed!

Good idea Scott. BTW it was in many of the photos posted earlier in this thread that we saw those HAWKs at intersections. MY guess based on personal experience with calling any traffic agency's attention to their not following the rec's or standards in the Manual is that you'll get a reply something like: Well the Manual was never intended to substitute for engineering judgment and we feel that based on site conditions, visibility etc, the current installation gives adequate service, bla bla bla, etc.

Lot's of fun. LOL........ BTW, the signal involved in that experience with my county DPW was finally changed a few years later.

[Scott5114]

Heh, yeah, that's a possibility. In that case you bring it up to an elected official or the media and shame them into compliance. "Engineering judgement blah blah blah" doesn't go very far when the six o'clock news goes demanding to talk to the engineer to get them to justify their "judgement", or when a state legislator asks for a meeting with the department head to explain why such an "unsafe" design was allowed...

Basically you contrive a situation in which doing the right thing (rather than nothing) is the path of least resistance.

[kphoger]

Heh, yeah, that's a possibility. In that case you bring it up to an elected official or the media and shame them into compliance. "Engineering judgement blah blah blah" doesn't go very far when the six o'clock news goes demanding to talk to the engineer to get them to justify their "judgement", or when a state legislator asks for a meeting with the department head to explain why such an "unsafe" design was allowed...

Basically you contrive a situation in which doing the right thing (rather than nothing) is the path of least resistance.

Especially if the non-compliant HAWK location is the site of a fatal accident.  Oh, wait, that isn't actually happening....

[mrsman]

There is a new HAWK signal in Los Angeles.

https://twitter.com/LADOTofficial/status/1542300646942167045

This is noteworthy, because as mentioned on this thread (and maybe the traffic signal thread), the old L.A. ped xing signal is far better than instituting a HAWK.  The old ped xing signals were NEVER placed at intersections, other than RIROs, in Los Angeles, and it seemed like Los Angeles was avoiding HAWKs at normal intersections as well.  As we discussed upthread, there are problems with HAWKs at intersections and unfortunately this feels like a major step backward for Los Angeles in particular.

[SignBridge]

It seems like more of these HAWKs are AT intersections than not, in conflict with MUTCD recommendations.

[mrsman]

It seems like more of these HAWKs are AT intersections than not, in conflict with MUTCD recommendations.

That seems to be my observation too.

FWIW, people are more likely to want to cross the street at intersections, rather than midblock.  That is because any pedestrian on the minor street would be added to those on the main street who may need to get over on the other side.  Another aspect is that bus stops tend to be on corners for similar reasons.  People who live on the side streets will walk to the main street to catch a bus, and they wouldn't unnecessarily want those people to walk an extra half-block.  Another issue is parking.  Most jurisdictions will have rules prohibiting parking near a crosswalk and near an intersection.  Overall, if all the crosswalks are at intersections, as opposed to mid-block, more on-street parking is preserved.

So more people are likely to want to cross the main street at an intersection with a small street, rather than mid-block.  But unfortunately, HAWKs are not as safe at intersections as they are at mid-block.  They are not meant to control cross-street movements, even though cross-street traffic will take advantage of the stopped traffic for the pedestrians.