POTHOLE ANALYTICS

Bad weather or bad management? What’s really driving New York City’s notorious pothole problem.

By Lucius Riccio

Pothole analytics

New York is famous for many things and many landmarks. From the Empire State Building and the Brooklyn Bridge to its ubiquitous taxicabs and glamorous nightlife, New York is world class. But there is one thing for which it should not be proud of and that is its large and numerous potholes. Everyone is aware of them even if they haven’t heard David Letterman’s regular jokes about them. (“There is a pothole so big on 8th Avenue it has its own Starbucks in it.”)

When it comes to potholes, some years seem to be worse than others. Without a doubt, this winter was an exceptionally bad year.

The city says it is proud of its pothole fixing efforts. City workers filled a record 300,000 potholes during the first four months of 2014. That’s an astounding accomplishment. But is it something to be proud of? Is filling hundreds of thousands of potholes emblematic of or reflective of good management?

Potholes are to a significant extent a measure of municipal competence and image. In addition, New York’s poor streets cost the average motorist an estimated $800 per year in repair work and new tires.

News reports and the statements of government officials have reinforced the idea that the (exclusive) reason for the avalanche of potholes and abysmal state of the city’s roads was the severe weather which brought bitter cold temperatures and a near record snow fall. Clearly that was a factor. But was it the major factor? Finding the answer was a job for analytics!

NYC’s Roadway System

About 30 percent of NYC’s land area is paved roadway. New York City has approximately 6,000 miles of streets as measured by the “white line in the middle of the street.” Since some streets are wider than others, a better measure of the city’s roadway system is the amount of lane-miles it has. The standard lane is 12 feet wide. On that basis, the city has about 20,000 lane-miles. That amount is almost the equivalent of a one-lane road around the world [x].

The city keeps good statistics on the condition of the streets and the number of potholes it fills. It reports those figures in the Mayor’s Management Report annually. It surveys the streets for their general condition and rates each street surveyed on a 10-point scale from 1 (poorest) to 10 (newly resurfaced.) It reports the percentage that is rated “good.” A rating of 7 or above is considered good.

It also surveys the streets regularly to determine where and how many potholes exist, and then sends those reports to its field crews. The crews then report their work. That information is also compiled and put in the Mayor’s report.

A Little Pothole History

 
Pothole analytics

Figure 1: Number of potholes fixed, 1996 to 2012.

 
Pothole analytics

Figure 2: Plot and linear fit of potholes vs. inches of snow.

 
Pothole analytics

Figure 3: Percent streets rated good.

 
Pothole analytics

Figure 4: Lane-miles resurfaced per year.

 
Pothole analytics

Figure 5: Trend of the gap.

 
Pothole analytics

Figure 6: Potholes vs. resurfacing gap.

Figure 1 shows a plot of the annual number of potholes fixed from1995 to 2013. 2014’s numbers are not in yet, but they will be more than 300,000. Motorists weren’t kidding when they said it was a terrible year.

Figure 1 shows the steady and dramatic increase in potholes from around 70,000 to 80,000 in the mid-1990s to the 200,000 to 300,000 range in the most recent years. Not only is the number in recent years astoundingly and devastatingly high, such a steep rate of increase has not been seen since the trauma of the 1970s fiscal crisis.

Using inches of snowfall as a measure of the severity of the winter, Figure 2 shows a plot of the number of potholes vs. the inches of snow each winter.

The hypothesis that the winter causes the potholes – that it is an act of God – has some merit. The relationship is significant, but not completely satisfying. Is there a more significant reason for the bumper crop of potholes?

Street Conditions and Road Repair

Back in the 1980s, using engineering principles, I, as NYC’s transportation commissioner, estimated that the city would need to resurface with a new coating of asphalt at least 1,000 lane-miles per year just to stay even with road deterioration. The 1,000 lane-miles would just keep the roads at whatever level they were at, good or bad, without getting better or worse. I used to call that “orbital velocity.” If the city did less than 1,000, the roads would get worse; more than 1,000, they would get better. To see real improvement, the city would have to do at least 1,200 lane-miles per year, and if it wanted to catch up with the backlog, at least 1,500 per year – “escape velocity.”

Figure 3 shows the streets had high ratings in the mid-1990s starting out with ratings around 85 percent good. The late 1990s and early 2000s saw a steady decline down to about 70 percent good. In the late 2000s the ratings held steady at around 70 percent, substantially worse than the beginning 85 percent but at least stable.

This is quite consistent with the estimates made in the 1980s. Figure 4 shows the lane-miles resurfaced per year. From 1996 through 2005, the amount of annual resurfacing was well below 1,000 (averaging less than 800), and the street ratings dropped consistently from the 85 percent range to the 70 percent range. As the amount of resurfacing approached and then achieved the 1,000 mark, the street conditions stabilized, albeit at a much lower level.

If the “1,000 lane-miles per year” figure is considered to be the “orbital velocity,” any amount below that would contribute to a “gap” or backlog of streets needing repair. The cumulative gap is displayed in Figure 5.

The current “gap” is close to 2,500 lane-miles, or about a one-lane road from New York to Phoenix! Can the gap predict potholes? Figure 6 shows the plot of potholes vs. the gap.

There is an indisputably strong relationship between the increase in the “gap” and the number of potholes. It is obvious that the real reason for the steady and substantial increase in the number of potholes is due to the increasing gap in road resurfacing.

Pothole analytics

Table 1: The data and regression analysis.

Pothole Analytics

Given that we have the data, it would be worthwhile to take advantage of all of it in making predictions about potholes and the implications of various policy choices. One way to do that is to perform a regression analysis using the resurfacing gap and inches of snow as the independent variables and number of potholes as the dependent variable.

That regression proved quite powerful. Its R squared (a measure of the goodness of fit) was .91, a very high number by statistical standards. The precise equation resulting from the regression was:

Potholes = 7801.5 + 80.6 *
Resurfacing Gap + 930.1 * Inches of Snow (eq. 1)

One way to interpret the equation is that there would be very few potholes (less than 8,000) if there happened to be no gap and no snow. For every lane-mile of gap, 80 potholes are produced, and for every inch of snow, 930 potholes are produced. With the continued gap of about 2,500 lane-miles, the model predicts the city will see at least 200,000 potholes next year plus an additional 45,000 to 50,000 potholes beyond that if the snow falls as it did this year. Table 5 lists the data for this analysis, the prediction from the regression analysis and the residuals (the difference from the actual and the predicted).

The correlations are clear. There is a 90 percent correlation between the number of potholes and the deficit in resurfacing. There is a 93 percent (negative) correlation between the deficit and the street ratings, and an 86 percent (negative) correlation between the street ratings and the number of potholes.

The pothole-creating effect of snow (and snow plowing) is likely significantly higher on older, worn out streets. As such if the ratings were to get even worse, the prediction provided by the regression model would be too low. Mathematically, more data may show a cross-product effect exacerbating an already bad problem.

Pothole Fixing: The Right Approach

Clearly, fixing potholes is an essential and commendable thing to do. And to do so efficiently is a worthy management objective. Of course, it is not how many you fill but how many you don’t fill. Or put another way, how long do they remain in the street breaking axles and blowing out tires? But in addition, I think the fixation (pun intended) with potholes is the wrong approach.

A high number of potholes is indicative of a failure to maintain the streets. Fixing potholes means the smart thing hasn’t been done, which is to do the work that prevents them in the first place. Potholes are emblematic of a failed strategy.

The point of this exercise was to demonstrate two things. First, analytics can help explain almost anything, including why there are so many potholes. I hope that serves as an inspiration to OR/MS people everywhere.

Second, America needs to take its infrastructure more seriously and develop clear and realistic plans for the maintenance of a state of good repair and for the future improvement of all of its infrastructure. Analytics can make a big contribution to that effort.

Lucius Riccio, Ph.D. P.E., is a senior lecturer at Columbia University’s Business School where he teaches business analytics, optimization and simulation and operations management. He also teaches in the School of Engineering, the School of International and Public Affairs and in the Sustainable Management program at the Earth Institute. He has been on the faculty of Columbia University since 1995. He also is executive vice president and partner of Gedeon GRC Consulting, a full-service engineering consulting firm.

Prior to joining the faculty at Columbia University, Riccio served in a variety of government, not-for-profit and private sector positions. Most notably, he was the Commissioner of NYC’s Department of Transportation and served on the Board of Directors of NY’s Metropolitan Transportation Authority. Prior to those positions he was the assistant commissioner of NYC’s Department of Sanitation, deputy commissioner for Highways (NYC), assistant director for research at the Police Foundation (Washington, D.C.), served on the staff of the President’s Commission on Law Enforcement Productivity, was a consultant to the Federal Judicial Center of the Supreme Court of the United States and the U.S. Department of Justice, and was senior vice president and partner at Urbitran Associates and managing director for special projects at Douglas Elliman Real Estate (NYC).

He has published extensively, served on the editorial board of the Operations Research journal and received the ORSA (now INFORMS) President’s Award for Lifetime Achievement.

References

  1. “Street Smart: A Plan to Improve and Equalize Street Smoothness Conditions in New York City,” Riccio et al., NYC DOT Report, 1988.
  2. Mayor’s Management Report, NYC Mayor’s Office of Operations, annually.
  3. National Weather Service Central Park Station.
  4. AAA Motorist, New York, N.Y.