Tag Archives: Safety

Roundabout Changes

I was a huge fan of the roundabout installation on Hillsborough St. near the NC State Bell Tower. It was well thought out, it reduced delay, and it was a pleasure to drive in. For a traffic engineer.

Unfortunately, everyone else had a collision in it. A little over a year ago, I personally wrote my recommendations here on my blog when numerous collisions were reported. I still stand by many of my points, especially the point on safety. No amount of fender-benders equal a fatality in my eyes, so I still say the intersection is safer than a four-way signalized intersection.

City and State engineers have done all they could to improve the design. Signage, pavement markings, flyers, flags, you name it. But downtown Raleigh drivers simply can’t afford the time to drive carefully in an unfamiliar design. With roundabouts being such an uncommon occurrence in North Carolina, I cede and will say that a two-lane roundabout of this size is too unfamiliar for uneducated drivers.

What changed my mind? I recently had lunch with Reza Jafari, President at Road Safety and Transportation Solutions, Inc. We talked at great length about driver education and the roundabout. He convinced me that the diameter is just too small for unfamiliar drivers, drivers that have never driven multi-lane roundabouts and are prone to change lanes (or disregard lane markings entirely) while navigating one. The point is fair.

We shall see how well a one-lane roundabout manages the traffic. Anything is better than a signal.

Reza Jafari is the President of Road Safety and Transportation Solutions, Inc. located in Cary, NC. He is a terrific resource on traffic safety and I would go to great lengths to recommend him and his company for a safety study.

Signal Warrants and Why We Use Them

A friend sent me a link to a recent WRAL article posted yesterday about a fatal collision in Cary on Sunday, February 12th1. Tragically, both passengers in a left-turning vehicle died when the driver crossed the path of another vehicle. This is terrible and my condolences go out to those affected by the incident. I also feel for safety engineers everywhere who feel the emptiness of failure at each reported fatality2, especially the highway safety professionals who work in this jurisdiction. However, as an engineer and scientist it is of utmost importance to step back and remember a few things about working within the confines of the real world: with limited tax funds, systems are imperfect and not every collision is preventable.

Keeping these things in mind, how do traffic engineers decide if an intersection requires a traffic signal? We use a system of “warrants”, or reasons that warrant the installation of a control device. The Manual on Uniform Traffic Control Devices (MUTCD) lists nine signal warrants for varying reasons in Chapter 4C. A pdf of this manual can be found here. These reasons range from volume to school-zone related conditions and yes, even crash experience.4 Most generally, traffic signals are never installed without first qualifying for one or more warrants.

Why? In many cases a signal can make an intersection worse than it started! Traffic signals are often viewed in the public eye as a cure-all, when in fact they can have adverse affects on safety and efficiency: increased delay, increased traffic control disobedience, increased use of other routes such as neighborhoods (just to skip the light!), and others. The increased stops can also increase the number of rear-end collisions at the site as well.4 This has led to the installation of many lights that were never needed, at sites that may now be less safe, at the cost of a six figure installation. Thats right, your average 4-way simple traffic signal costs between $150,000-200,000. Plus continued maintenance from now until forever.

The Media

One of the biggest reasons that the public is so ill informed about the affects of traffic signals is the media, and we can see it in this article.2 It’s not necessarily their fault that the public demands stories that move the heart, either in tenderness, controversy, or horror. But people get so caught up in political partisan biases, that many forget how easy it is to be biased to the heart instead of the mind, allowing an element of “controversy” to develop conspiracy, scheme, even frustration in the reader. Lets take a moment to analyze some of the diction used in this article:

“hasn’t met the criteria for a traffic light, despite pleas from drivers and neighbors that the intersection is too dangerous”

I’d be pulling my hair out if there was any left to pull. On the logic side, uninformed readers could assume two very incorrect conclusions: drivers and neighbors always know a dangerous intersection when they see one3, and that pleas from drivers and neighbors warrant a traffic signal. But theres so much on the emotional side as well. Using the word “plea” to stab the hearts of our readers, the author has clearly set the tone of the entire article to:

“People have begged on hand and knee, offering to sacrifice their first born for just one, ONE traffic signal. The site would be safer, and this wouldn’t have happened”.

No wonder people think traffic signals are a cure-all! This conjecture could be patently false. For all we know without adequate evidence, adding a traffic signal could make this site WORSE. But not now. As uninformed citizens we now know that the evil city of Cary is denying us the promise of safety and security that only a traffic signal can bring.

“she attempted to turn left from West Chatham Street onto Cary Parkway and crossed into the path of a Dodge Avenger.”

This one earns the reporter kudos. Notice how quickly and easily the blame of the incident could have been shifted: 1) “she unwittingly pulled in front of traffic”, or 2) “the Dodge Avenger smashed into the side of her vehicle”. Unfortunately for the reader, we don’t know the exact cause of the collision. We may never know for sure. It could’ve easily been texting or another form of distracted driving. The author calls this collision “correctible”, though as a reader we are left unsure as to whether or not it is she or her interviewee that thinks it’s correctible and why. But by calling this collision “correctible”, the public will be wondering why correctible collisions are happening at all. And they should! But they are missing an important piece of the puzzle: some intersections are more correctible than others and warrant more immediate attention and funding, of which there is a finite amount of both.

“In addition to traffic volume and other criteria, to warrant a light an intersection would need to have had five crashes that would have likely been corrected by a traffic light in a single year.”

But I have to draw the line here, this is a complete falsehood and is misreported. Intersections do not require crashes for a traffic signal. In fact, new roads are constantly being built in conjuction with signals before they are even opened. But the remark is so instigating and confusing, it left one commenter with a terrible impression of traffic safety methods. If anything should make an engineer feel like a failure, it is the terrible miscommunication with the public that can lead to this:

Let me see if I have this right….The criteria for installing a traffic light requires multiple occurrences of either property damage, bodily harm, or loss of life. The wisdom of our city fathers leaves much to be desired. carrydoggymom, February 13, 2012 7:55 p.m

Concerned Citizens

As much as a I try to avoid reading the comments section on a news website, it’s important from a public administration point-of-view to see how this fraction of the populace thinks. Some comments are enlightening, some are sad. This time, I was mostly surprised to see intelligent debate in the comments section. Mostly.

“I bet if it was the mayor’s family, there would be a light there tomorrow. What a load of bull!” bjandroxie88, February 13, 2012 7:42 p.m

Here is the advocation that, not only should tax dollars be spent willy-nilly by those in power when they find misfortune, but that politicians should make engineering decisions. There is a problem when the politically inclined try to mingle in the affairs of those who put safety first, a perfect reason why engineers should use MUTCD warrants instead of bending to political pressure from above.

“…If the vehicle was close enough to hit them, it was close enough to see and avoid.” pedsrndad, February 14, 2012 11:56 a.m

While my first instinct is inclined to agree, sight distance plays an important factor in intersection design, and it may need reinvestigation at this particular intersection. Though I imagine it was addressed and not seen as a big problem during the original design, factors such as reclassifying the speed-limit or increased volume year-to-year could play a part in changing the environment here.

The Intersection

I’ve driven the intersection myself, and especially during rush hour it can be difficult to turn left. The median supplies a little refuge, but not much. I glanced at the traffic volume maps, but without knowing more about peak hour data, it’s difficult to say exactly what warrants the intersection can qualify for. I’m surprised it doesn’t hit the peak-hour warrant, or even the school-crossing warrant with it’s proximity to Laurel Park Elementary. With the limited data I have on hand, I might recommend the following traffic studies to take another look at the signal warrants: sight distance study, spot speed study, and/or a volume study.5

If a light is installed in the near future, I hope it will be because it is warranted and that conditions have changed since the last inspection of the intersection. Not because of the ravings of uninformed watchers of the 6 o’clock news. We should all strive to be more cognizant of the impact media can have on the heart of the public, and how important good communication can be when working for the public sector. These efforts will make it easier to limit the affect of a tragic story in the news from opening our wallets too wide when there aren’t enough collisions to merit a new, expensive, and possibly ineffective traffic control device.

Wrap Up (Read: Important Conclusions)

By now you should have gathered the following:

  • Traffic signals are not a cure-all; sometimes, when not really needed, they cause problems.
  • We use Signal Warrants to come to justified conclusions, recommended by the MUTCD, on whether or not to add a traffic signal at an unsignalized intersection.
  • MUTCD has 9 different signal warrants for varying reasons of safety.
  • Intersections do not require collisions to warrant a traffic signal.
  • “Pleas from drivers and neighbors” is not an MUTCD signal warrant.
  • “The mayor found misfortune” is not an MUTCD signal warrant.
  • Crashes and fatalities do not necessarily merit an MUTCD signal warrant.

References and Notes:

  1. Cary intersection doesn’t meet criteria for traffic light, WRAL, Monday, February 13th, 2012
  2. Mental Note: Prevented collisions never make the news. How could they?
  3. Mental Note: Drivers and neighbors are an invaluable resource when discussing what they think of an intersection. However, everyone’s seen a collision somewhere. This does not always, a “dangerous” location, make.
  4. Manual on Uniform Traffic Control Devices 2009, published by the Federal Highway Administration.
  5. As found in the Highway Safety Engineering Studies Procedural Guide, published by USDOT and FHWA 1991, commonly referred to as “The Parker Manual”.

The Lateral Forces of Earthquakes

So it seems there were a few damages from the Virginia earthquake earlier this week. The biggest damages on the news? The National Cathedral and the Washington Monument. The Washington Monument suffered some cracking at the very top and has since been closed indefinitely to the public until damages can be assessed. As for the National Cathedral, gargoyles, spires, buttresses, and walls cracked, shifted, or fell and shattered. Other homes and businesses near the epicenter were damaged as well.

While damages under 6.0 earthquakes are rare, they do indeed occur, especially in a region less known and under-designed for quakes. In fact, the area falls in a zone of very small seismic risk (see the 2012 International Building Code Map, courtesy of USGS), which is a big part of the problem. Another part of the problem is the age of these structures. Newer buildings and building methods are much safer than they used to be in this regard, but many older buildings will suffer problems, especially since there is little desire to improve them and little funding to do so.


Figure 1: Building Motion During an Earthquake (1). Click to Enlarge

What causes these damages? The lateral forces caused by the accelerating displacement of the ground. When an earthquake occurs, the acceleration of the ground will cause the building to move sideways at the base of the building. It is, after all, firmly attached to the ground in most cases, causing a lateral load and an equivalent shear force at the base (see Figure 1). The building will then begin to swing to and fro, according to the change in direction of the ground.

Familiar with Newton’s law of force? You could very simplistically apply it here to calculate the forces on the building: F = M(A), or force equals the mass multiplied by the acceleration. What may not be immediately obvious is that the higher the mass of the building, the higher the force on the building. But we CAN control the acceleration, which gives us an advantage if we play our cards right. The acceleration can be affected by the natural period of the building, or a complete oscillation, which is dependent on the building stiffness.

Supposing we could create a building that was perfectly stiff, it would match the acceleration of the ground perfectly and it would not oscillate and not experience force. This, unfortunately, is an impossibility for any material, and the slightest deformation would cause large forces due to short natural periods. So contrary to instinct, we do not want to make the stiffest buildings possible, what we want are flexible, long natural periods in our buildings. We need them to sway.


Steel buildings have certainly come a long way, but a large part of the problem with older buildings is their dependence on stone, under-reinforced concrete, unreinforced masonry, and designs incapable of holding even the most moderate earthquake and wind loads. Many of these buildings were built at the turn of the 20th century and are still being used today.

Stone and concrete hold well under compression, but not under tension. Without an element of tension, these older stone buildings simply crumble. Newer concrete structures today are reinforced with steel, doing wonders for flexibility. When allowed to, the steel in reinforced concrete catches tension loads and transfers them. Steel is ductile, and the greater the ductility of a building, the better forces can be absorbed. When designed correctly, even after structural failure, warning signs in reinforced concrete structures are easily apparent, often allowing people to clear from the site before catastrophic failure. After all, people are more important than the buildings.

Figure 2: Deformation Components of a Reinforced Concrete Column (2). Click to Enlarge

While working on a graduate research project during my time at the Constructed Facilities Lab at NC State, I worked on a thesis by Pablo Robalino (2). We tested lightweight concrete columns for seismic lateral forces to see flexural and shear deformation. Notice in Figure 2 how the flexural damage occurs on the side of the lateral force, while the shear damage occurs throughout the column towards the ground. The combination of these effects must be considered when designing columns for the lateral loads associated with earthquakes.

Figure 3: Column Specimen From Seismic Test

Figure 3 is a chunk of a column we tested. Notice the lines drawn with permanent marker. These lines follow cracks in the concrete caused by forces on the column. As we tracked their progression, they naturally followed similar trajectories as those depicted in Figure 2.

These aren’t difficult concepts to understand, but even with the best methods of absorbing forces, costs often limit the investment we can place in a structure to prepare for the worst. While we can’t feasibly prepare for earthquakes of every magnitude, we can use physics, properties of materials, and our growing understanding of these natural disasters to build structures capable of sustaining many of the forces that seem beyond our control.


  1. Professional Publications Inc. “Lateral Forces – Earthquakes”
  2. Robalino, Pablo. “Shear Performance of Reinforced Lightweight Concrete Square Columns in Seismic Regions”. August, 2006.

Railroad Grade-Crossing Hazards

Two weeks ago, an unfortunate collision occurred at a railroad grade crossing in Maine. Reuters reports that a dump truck was hit by an Amtrak train and the driver was fatally wounded. Four passengers aboard the train were injured as well. Terrible. At-grade railroad crossings are some of the most dangerous intersections we have on our road system and they should be avoided whenever possible. Fortunately, many public agencies are fully aware of hazards associated with them and are taking steps to fix them. Here is a list of policies in the United States associated with at-grade crossings, published by the Federal Highway Administration. If you live in North Carolina, rest assured that NCDOT can and will use its power to remove, abandon, close, or regulate all railroad grade crossings. That is, if politicians don’t try to stop the sensibilities of the engineers there.

You might also be interested to know that railroad companies are also very aware of these hazards. CSX, for example, is “firmly opposed” to at-grade crossings and supports policies in place by the USDOT and state agencies to limit their use. You can read more about their leadership in this area on their website.

It’s a shame to see collisions on train tracks. Look both ways when you have to cross tracks and don’t, under any circumstances, try to “beat” the train or drive around protective barriers. It’s not worth it.

Full story at Gizmodo.

What You Can Do About NC-H.561

The following is an email I sent to Wake County Representatives, concerning the issue of legislation-run engineering. If you can, write something similar RIGHT NOW to your representatives. A list of NC Representatives can be found here. Here is a list of NC Senators.

From: Mike Roselli 
To: Jennifer.Weiss@ncleg.net, Paul.Stam@ncleg.net,
Deborah.Ross@ncleg.net, Tom.Murry@ncleg.net,
Grier.Martin@ncleg.net, Darren.Jackson@ncleg.net,
Rosa.Gill@ncleg.net, Nelson.Dollar@ncleg.net,
Marilyn.Avila@ncleg.net, Richard.Stevens@ncleg.net,
Josh.Stein@ncleg.net, Neal.Hunt@ncleg.net,

Senators and Representatives,

I live in Wake County and I work in the engineering field. I’ve been made aware of H561, a bill to supersede engineering safety design decisions made by engineers at NCDOT. I urge you to NOT back this bill. It is not based on facts, research, or scientific credibility, but on fears of change. It has been well established in the engineering community that highway medians substantially increase driver safety and save many lives every year without impacting local businesses.

A research paper on this topic can be found here: http://blogs.newsobserver.com/sites/drupalblogs.newsobserver.com/files/docs/ncsu2009-12FinalReport.pdf

Over 30,000 people will die this year in the U.S. from motor vehicle collisions.[1] We cannot set a precedence of State legislature overseeing every engineering decision made in the name of public safety. Do we remove stop signs when businesses ask? Do we remove bridge supports when someone finds them unappealing? That is, in essence, what some are suggesting: taking the complaints of the uneducated few and using these complaints to stifle progress in public safety and safety awareness at the cost of human lives.

I have a write up on my website with more details.

Thanks for everything you do,
Mike Roselli
Wake Co Resident

1. Source: List of Motor Vehicle Deaths by Year

Mike Roselli is a graduate of the NC State Civil Engineering Department. This writing reflects the views, opinions, and judgement of the author and does not necessarily reflect the views, opinions, or judgement of NC State University or NCDOT.

Open Letter to Rep. LaRoque

The following is an email I sent to Representative Stephen LaRoque, concerning the issue of legislation-run engineering:

From: Mike Roselli 
To: Stephen.LaRoque@ncleg.net
Sent: 5/11/2011 12:42pm

Dear Representative LaRoque,

I work for NCDOT and I have studied Highway Safety at NC State. Not only do I find Bill 561 a despicable act of politics, but I find it offensive for you to imply that you know more about the engineering of highways than the engineers at NCDOT. I urge you to reconsider your endorsement of this bill and read the facts. Here is a fantastic report on how medians do NOT affect local businesses, studied by researchers at NC State and UNC:


It will be a sad day when State engineers, who work tirelessly to provide life-saving solutions for collision problems, have to report to the state legislature about every median, stop sign, or other collision-reducing counter measure they want to install.

Mike Roselli

Politics “Bypassing” Engineering

Today, the News and Observer reports that the State Senate will be voting on House Bill 561, to require NCDOT to remove a half mile of center median installed on new projects in Asheville, NC, “at a cost estimated by NCDOT at $797,500”, and to give the State legislature power over road design for other projects.

Other than ripping up a nearly completed project to nullify the design, the legislation would force DOT to change widening plans and forget other median installations planned on other routes. The fact that politicians, who know next to nothing about road design, can be influenced by their constituents to change engineered designs is unfathomably horrific.


Restricting left turns is good for safety. Median installations that require drivers to plan their left turns and u-turns at a limited number of intersections increase road safety by decreasing all sorts of collisions and allowing traffic to flow more smoothly. The decisions made in these designs were based on safety.

So who has a problem with this? Well, access restriction is seen by many locals as harmful to businesses, despite some studies showing evidence to the contrary. The latest is by a research team here at NC State and UNC, titled “Economic Effects of Access Management Techniques in North Carolina”. I’ve attached the study and you can download it here.

The study confirms that little to no evidence of economic problems was found after the installation of turn-restricting medians.

Additionally, it’s important to note that drivers are more aware of road safety than they think. And in fact, some drivers may be more inclined to use your business if they feel safer about entering and exiting the access to it. The Federal Highway Administration has a great write up on safe access and what it means for businesses. I’ve attached it and you can read it here.

The Problem

Indeed, with enough fussing to the legislator whose campaign you backed, it seems businesses may be able to “buy” the road design they want for their businesses, regardless of safety concerns. Even in the face of state budgetary concerns, it doesn’t seem to bother Rep. Stephen A. LaRoque to use another million dollars to remove a brand new installation developed in the interest of public safety. What exactly seems to be going through his mind? This:

“The legislature should have the final say on projects like this before they go through.” -Rep. Stephen A. LaRoque

This is a problem. It is a sad day when engineers who make sound decisions based on data can have their life-saving designs overseen and rejected by bought elected officials in the name of cheap tricks and politics.

Bottom line? Let engineers do their jobs. And write your representatives an email and tell them. I did.

Source: News & Observer


It seems the Senate has delayed voting on this issue for now. I’ll be posting more on this as it develops.

New Roundabout, New Collisions

I was reading this News & Observer article today about the new roundabout on Hillsborough St in Raleigh, NC. Raleigh Police have cataloged more than 40 collisions at a new roundabout installation at it seems at least a few people have some ruffled feathers.

At first this number seems high, but it’s important to remember a few key thoughts about the safety, design and installation of new traffic patterns:

New patterns can cause collisions

It’s a fact, and one that is difficult to adjust for. New traffic patterns will confuse motorists and can cause collisions. It’s to be expected. While forty seems like a rather high number, high rates during a learning period are why safety engineers do not normally begin their investigation for at least 12 months after a new traffic pattern.

Reduction in conflict points

Roundabouts drastically reduce the conflict points at an intersection. While less conflict points do not necessarily reduce collisions, there are less places vehicles can collide, one of the large advantages of roundabouts.

Reduction in conflict speed

Roundabouts also reduce the speed of incoming traffic. When drivers approach a roundabout, they approach a yield as well as a significant turning radius to enter the circle. This reduction in speed reduces the severity of collisions in the roundabout. The magic of the roundabout is that all of this is done while still reducing delays and increasing the flow of traffic!

Collision severity tradeoffs

When evaulating safety, the number of collisions is not the only metric analysts are concerned about; severity is a metric which should be considered at great length. Without looking at detailed traffic collision reports for the area, it is impossible to determine the safety of the previous intersection. However, it should be noted that roundabouts do not normally have head-on collisions. Because of this, almost all collisions at roundabouts will be of a much lower severity than at a standard 4-way signalized intersection. In fact, only one of the over 40 reported collisions had a severe injury and it involved a motorcycle.

Of course, you should be comparing to injuries and fatalities with the previous design. But how many fender benders are equivalent to severe injuries? Or fatalities? It’s difficult to draw a line here. The following is known by many as “The Old Kentucky Formula”, an easy to use equation for an Equivalent Property Damage Only index (EPDO) based on the data provided by a field officer on the scene of a reported collision:

EPDO = 9.5 (F + A) + 3.5 (B + C) + PDO


  • FFatality: Collisions resulting in one or more.
  • AAmbulence: Injury severe enough for an ambulence to be called to the scene.
  • BBruise: A visible, but non-emergecny injury
  • CComplaint: Officer cannot visably see the reported injury.
  • PDOProperty Damage Only: No persons report an injury on scene.

The formula is an older one so many state agencies have developed their own. NCDOT has their own Severity Method treatment which can be found here.

Final Thoughts

A city safety analyst seems to be keeping tabs and will conduct an investigation after the roundabout has been in use for a full year. I think they should throw out, at the very least, the first six months of data plus the months of construction and signage installation time. The study may very well show areas that can be improved such as driver education, signage, pavement markings, lighting, etc. But judging from my own paths through the intersection and the severity of reported collisions mentioned in the article, there is no doubt in my mind that the roundabout was the correct treatment for the intersection, and while a few comments on the article from readers show a few disgruntled folks, many more applaud it for it’s efficiency. Give it some time and the frequency of the fender benders will reduce.

Mike Roselli is a graduate of the NC State University Civil Engineering Department. He works for the North Carolina Department of Transportation. The views expressed in this post reflect the judgement of the author and do not necessarily reflect the views, opinions, or judgement of NC State or NCDOT.

Safety Study

For my Highway Safety graduate level course at NC State, I conducted a field study at the intersection of Glenwood Ave and Peace St in Raleigh, NC. The study evaluated the current status of field conditions, conflicts, previous collision reports, and recommended a few counter measures that might reduce conflicts at the intersection.

I’m really enjoying the class. If you’re interested in reading the study, I’ve attached it to this post.

Safety Report