green wave for walking?: Indeed, as the urbanist William H. Whyte pointed out, the signals on Fifth Avenue seem designed to thwart the pedestrian: “Traffic signals are a particular vexation. They are, for one thing, timed to benefit cars rather than pedestrians. Take Fifth Avenue. You want to make time going north. At the turn of the light to green you start walking briskly. You have about 240 feet to go to reach the next light. You will reach it just as the light turns red. Only by going at flank speed, say 310 feet per minute, will you beat the light.” From William H. Whyte, City (New York: Doubleday, 1988), p. 61.
even higher authorities: This is not such a far-fetched premise. A study by a team of researchers at Bar-Ilan University in Israel examined pedestrian behavior in two cities: the “ultra-Orthodox” Bnei Brak and the “secular” Ramat-Gan. While traffic and infrastructure conditions were essentially the same in both locations, pedestrians in Bnei Brak were three times more likely to commit what the researchers judged “unsafe” pedestrian behaviors. This may be a function of the fact that fewer residents of Bnei Brak own cars; thus they’re less cognizant of drivers’ abilities or less willing to consider them. But the researchers suggested another reason, citing studies that note “a strong connection between the belief in supremacy of other laws (i.e. religious laws) over state laws, and a readiness to violate the law.” See Tova Rosenbloom, Dan Nemrodova, and Hadar Barkana, “For Heaven’s Sake Follow the Rules: Pedestrians’ Behavior in an Ultra-Orthodox and a Non-Orthodox City,” Transportation Research Part F: Traffic Psychology and Behaviour, vol. 7, no. 6 (November 2004), pp. 395–404. For more on the link between religious belief and compliance with laws, see A. Ratner, D. Yagil, and A. Pedahzur, “Not Bound by the Law: Legal Disobedience in Israeli Society,” Behavioral Sciences and the Law, vol. 19 (2001), pp. 265–83.
“crosswalk on the Sabbath”: Letter from the Rabbinical Council of California to John Fisher, August 9, 2004.
stops by 31 percent: F. Banerjee, “Preliminary Evaluation Study of Adaptive Traffic Control System (ATCS),” City of Los Angeles Department of Transportation, July 2001.
previous night’s fireworks: In 2005, the CHP reported, there were thirty-four Code 1125-A incidents on Tuesday, July 5, roughly 50 percent more than the previous or following Tuesday. Data provided by Joe Zizi of the CHP.
“driving on ice, literally”: The link between precipitation intervals and crash risk is well-known driver lore, and studies back it up. See Daniel Eisenberg, “The Mixed Effects of Precipitation on Traffic Crashes,” Accident Analysis & Prevention, vol. 36 (2004), pp. 637–47.
for many decades: G. F. Newell, a researcher at the University of California at Berkeley, observed that “in later years, indeed even to the present time, some researchers try to associate with vehicular traffic all sorts of phantom phenomena analogous to the effects in gases. They don’t exist.” He also argued that traffic is not “like any of the idealized models that the mathematical statisticians theorize about. It is messy and can be analyzed only by crude approximations.” G. F. Newell, “Memoirs on Highway Traffic Flow Theory in the 1950s,” Operations Research, vol. 50, no. 1 (January–February 2002), pp. 173–78.
“puzzles remain unsolved”: See Carlos Daganzo, “A Behavioral Theory of Multi-lane Traffic Flow, Part I: Long Homogeneous Freeway Sections,” Transportation Research Part B: Methodological, vol. 36, no. 2 (February 2002), pp. 131–58.
“heterogeneity of driver behavior”: In his superb book Critical Mass, Philip Ball, noting the increasing inclusion of “psychological” and other such factors in traffic modeling, points out a conundrum: “The more complex the model, the harder it becomes to know what outcomes are in any sense ‘fundamental’ aspects of traffic flow and which follow from the details of the rules.” See Philip Ball, Critical Mass (New York: Farrar, Straus and Giroux, 2004), p. 160.
when they followed passenger cars: The researchers who conducted the study speculated that following drivers may believe that SUVs, like tractor-trailers, take longer to stop than a car, and thus it is safer to follow at a closer distance. Another theory is that “ignorance is bliss”—that is, drivers worry less about what they cannot see than what they can (or they merely focus on the vehicle immediately in front of them, rather than a stream of several vehicles, because it seems easier). See James R. Sayer, Mary Lynn Mefford, and Ritchie W. Huang, “The Effects of Lead-Vehicle Size on Driver-Following Behavior: Is Ignorance Truly Bliss?” Report No. UMTRI-2000-15, University of Michigan, Transportation Research Institute, June 2000.
Los Gatos effect: Carlos F. Daganzo, “A Behavioral Theory of Multi-Lane Traffic Flow,” Part I: Long Homogeneous Freeway Sections.” Transportation Research Part B: Methodological, vol. 36, no 2 (Febryary 2002), pp 131–58.
traveling at 55 miles per hour: In 1985, the Highway Capacity Manual, the bible of highway engineers, put maximum capacity at 2,000 vehicles per lane per hour. That was raised to 2,300 in 1994 and raised again in 1998 to its current figure. Drivers, it seems, are willing to drive at a closer distance to the car ahead of them and to do so at higher speeds in the past. Why are drivers willing to take on more risk? It may be because vehicles have better handling, or because drivers are finding themselves having to cover more distance in a commute, and are thus willing to drive more aggressively to reduce the time. See Federal Highway Administration, 2004 Status of the Nation’s Highways, Bridges and Transit: Conditions and Performance (Washington, D.C.: 2004), U.S. Department of Transportation, pp. 4–16. Similarly, where previous estimates calculated that maximum flow occurred at 45 miles per hour, research by Pravin Varaiya in California, drawn from inductor-loop figures, now puts that figure at 60 miles per hour. See Z. Jia, P. Varaiya, C. Chen, K. Petty, and A. Skabardonis, “Maximum Throughput in L.A. Freeways Occurs at 60 MPH,” University of California, Berkeley, PeMS Development Group, January 16, 2001.
that it is being underused: As with many things in traffic, there is a debate as to the actual efficacy of HOV lanes from a traffic point of view (and not a social perspective). Do they improve the total flow of the highway or, more narrowly, simply give HOV drivers a faster trip? Or do they actually accomplish neither? In one study, by University of California researchers Pravin Varaiya and Jaim-young Kwon, based on loop-detector data taken from freeways in the San Francisco area, the HOV lane, it was argued, not only increased congestion in the other lanes (as one might expect if only a minority of drivers are using the HOV lane), but itself suffered from a 20 percent “capacity penalty.” The reason? As it was a single lane, any driver stuck behind a “snail”—in California, driving 60 miles per hour earns you this characterization—in the HOV lane had to travel the speed of the snail (as the other lanes were even slower, it would not do to try to pass the HOV snail). An additional potential complication that has emerged is that in California, cars bearing a hybrid fuel sticker (85,000 of the most recent version were issued) are legally permitted to drive in HOV lanes. Those drivers may indeed wish to travel around 60 miles per hour, as that will produce higher fuel efficiency (as indicated by the in-car displays). In a later study by fellow University of California researchers Michael J. Cassidy, Carlos F. Daganzo, Kitae Jang, and Koohong Chung (of the California Department of Transportation), the authors reexamined Varaiya and Kwon’s data and came to the conclusion that while overall traffic speeds did drop concurrently with the time the HOV lane was actuated (which, it must be pointed out, is precisely when the roads begin to get crowded; hence the HOV lane), they could not attribute this decline to the HOV lanes themselves, and in some cases, the HOV lanes actually enhanced the flow of traffic through troublesome bottlenecks. See J. Kwon and P. Varaiya, “Effectiveness of High-Occupancy Vehicle (HOV) Lanes in the San Francisco Bay Area,” July 2006, available at http://www.sci.csuhayward.edu/~jkwon/, and Michael J. Cassidy, Carlos F. Daganzo, Kitae Jung, and Koohong Chung, “Empirical Reassessment of Traffic Operations: Freeway Bottlenecks and the Case for HOV Lanes,” Research Report UCB-ITS-RR-2006-6, December 2006.
<
br /> nowhere near critical density: “Possible Explanations of Phase Transitions in Highway Traffic,” C. F. Daganzo, M. J. Cassidy, and R. L. Bertini, Department of Civil and Environmental Engineering and Institute of Transportation Studies, University of California, Berkeley, May 25, 1998.
If done properly: This is not to say that ramp meters always work perfectly, because nothing in traffic is ever so easy. Timing patterns may be skewed (although this is being addressed with real-time, system-wide adaptive ramp meters). Ramp metering done without carefully studying the traffic terrain can lead to “perverse outcomes,” one study suggests, as in the case of metered on-ramp drivers being held hostage by a “downstream” off-ramp they will not even use (congestion caused “not by too many cars getting on the freeway but by too many cars trying to get off”). Too many cars held on the ramp, no matter how desirable for the freeway, can back up into local streets, triggering other jams. Needless to say, for metering to work properly, people actually need to obey the signals. There is a fairness issue as well, as the authors of the Minnesota study pointed out: Ramp metering favors those making longer trips and actually hurts those traveling only a few exits. See Michael Cassidy, “Complications at Off-Ramps,” Access magazine, January 2003, pp. 27–31.
one-third less time: The rice experiment (proposed by Paul Haase) was the winning entry in a contest sponsored by the Washington DOT for the best way to visualize “throughput maximization” Susan Gilmore, “Rice Is Nice When Trying to Visualize Highway Traffic,” Seattle Times, December 29, 2006.
“like cars on the highway”: To wit: “Traffic flow resembles granular flow nowhere more closely than on the highway. Here the individual behavior of the drivers forms a relatively small statistical perturbation on the deterministic part of the collective motion, and hence the cars can be treated as physical particles. Both are many particle systems far from equilibrium, in which the constant competition between driving forces and dissipative interactions leads to self-organized structures: Indeed, there is a strong analogy between the formation of traffic jams on the highway and the formation of particle clusters in a granular gas.” From K. van der Weele, W. Spit, T. Mekkes, and D. van der Meer, “From Granular Flux Model to Traffic Flow Description,” in Traffic and Granular Flow 2003, eds. S. P. Hoogendoorn, S. Luding, P. H. L. Bovy, M. Schreckenberg, and D. E. Wolf (Berlin: Springer, 2005), pp. 569–78. On the other hand, G. F. Newell, a seminal traffic flow researcher, once cautioned that “some researchers try to associate with vehicular traffic all sorts of phantom phenomena analogous to the effects in gases. They don’t exist.” G. F. Newell, “Memoirs on Highway Traffic Flow Theory in the 1950s,” Operations Research, vol. 50, no. 1 (January–February 2002), pp. 173–78.
“through the hopper”: Rice is not a perfect metaphor for traffic either. As Benjamin Coifman points out, “Traffic is mostly a one-dimensional system within the lane, with occasional coupling to adjacent lanes. Traditional granular flow is three-dimensional. And then in traffic you are dealing with smart particles.” (Author interview.)
between the grains: The German physicist and traffic researcher Dirk Helbing has observed a similar phenomenon at work in the “outflow” of people from crowded rooms. “Panicking pedestrians often come so close to each other, that their physical contacts lead to the buildup of pressure and obstructing friction effects.” This can occur even when the exits are fairly wide. Why? “This comes from disturbances due to pedestrians, who expand in the wide area because of their repulsive interactions or try to overtake one another.” His simulations have found that columns placed asymmetrically in front of door openings can help “reduce the pressure at the door.” As with rice, when you organize the flow, slower is faster. See Dirk Helbing, “Traffic and Related Self-Driven Many-Particle Systems,” Reviews of Modern Physics, vol. 73, no. 4 (2001), pp. 1067–1141.
with ramp meters than without: See David Levinson and Lei Zhang, “Ramp Meters on Trial: Evidence from the Twin Cities Metering Holiday,” Department of Civil Engineering, University of Minnesota, May 30, 2002; see also Cambridge Systematics, “Twin Cities Ramp Meter Evaluation,” prepared for Minnesota Department of Transportation, February 1, 2001.
rarely have to stop: Jerry Champa, “Roundabout Intersections: How Slower Can Be Faster,” California Department of Transportation Journal, vol. 2 (May–June 2002), pp. 42–47.
1,320 vehicles per hour: Robert Herman and Keith Gardels, “Vehicular Traffic Flow,” Scientific American, vol. 209, no. 8 (December 1963).
more lost time: According to one study, SUVs reduce traffic flow in another way as well, by blocking the view of following drivers, who tend to leave more headway as their sight distance drops and they are less sure of traffic conditions ahead. This, of course, diverges from the findings of another study, cited above in the note for the phrase “when they followed passenger cars.” The difference in results may be due to the different types of roads on which the two studies were conducted or some other unidentified artifact. Kara M. Kockelman and Raheel A. Shabih, “Effect of Vehicle Type on the Capacity of Signalized Intersections: The Case of Light-Duty Trucks,” Journal of Transportation Engineering, vol. 126, no. 6 (1999), pp. 506–12.
stop on red: See, for example, Matt Helms, “Wait Just Two Seconds Before You Start,” Free Press, June 18, 2007.
Drivers talking on cell phones: University of Utah psychology professor David Strayer found in one driving-simulator experiment that subjects talking on a cell phone tended to drive more slowly and make fewer lane changes to avoid slower moving traffic (which may be read as a surrogate for a delayed ability to react). The total of this activity, Strayer estimates, adds 5 to 10 percent to total commuting times (then again, driving more slowly has safety and environmental benefits). See Joel M. Cooper, Ivana Vladisavljevic, David L. Strayer, and Peter T. Martin, “Drivers’ Lane-Changing Behavior While Conversing on Cell Phone in Variable-Density Simulated Highway Environment,” paper submitted to 87th Transportation Research Board meeting, Washington, D.C., 2008.
about 12 miles per hour: Robert L. Bertini and Monica T. Leal, “Empirical Study of Traffic Features at a Freeway Lane Drop,” Journal of Transportation Engineering, vol. 131, no. 6 (2005), pp. 397–407.
wreak progressive havoc: See Philip Ball, “Slow, Slow, Quick, Quick, Slow,” Nature, April 17, 2000. For the original research, see T. Nagatani, “Traffic Jams Induced by Fluctuation of a Leading Car,” Physical Review E, vol. 61 (2000), pp. 3534–40.
effects of a shock wave: See P. Breton, A. Hegy, B. De Schutter, and H. Hellendoorn, “Shock Wave Elimination/Reduction by Optimal Coordination of Variable Speed Limits,” Proceedings of the IEEE Fifth International Conference on Intelligent Transportation Systems (ITSC ’02), Singapore, pp. 225–30, September 2002.
trip times declined: Highways Agency, M25 Controlled Motorways: Summary Report, November, 2004.
slower can be faster: These systems require careful planning, however, to avoid unintended effects. The speed-limit step-down cannot be too sudden, or that itself could cause a shock wave. The ideal system would be coordinated along the length of the highway, to avoid simply sending one well-coordinated group of drivers smack into another jam farther down the road—and inadvertently helping to extend that jam or cause another one. See, for example, P. Breton et al., “Shock Wave Elimination/Reduction by Optimal Coordination of Variable Speed Limits.”
or the opposite: Boris Kerner notes, “The traffic flow instability is related to a finite reaction time of drivers. This reaction time is responsible for the vehicle over-deceleration effect: if the preceding vehicle begins to decelerate unexpectedly, a driver decelerates stronger than is needed to avoid collisions.” From Boris Kerner, The Physics of Traffic: Empirical Freeway Pattern Features, Engineering Applications, and Theory (Berlin: Springer, 2004), p. 69.
each car behind it will stop: One simulation compared the “oscillations” and “amplifications” found in stop-and-go traffic to those found in queues. “Perturbations” i
n the queue, or the way people stopped and started, were often observed to grow larger from the front to the back of the queue in simulators using cellular automata. See Bongsoo Son, Tawan Kim, and Yongjae Lee, “A Simulation Model of Congested Traffic in the Waiting Line,” Computational Science and Its Applications: ICCSA 2005, vol. 3481 (2005), pp. 863–69.
the harder it is to predict: An interesting parallel has been drawn between the way nonlinear traffic flows behave and the way supply chains work in the world of business. Supply chains suffer from what has been called the “bullwhip effect”—the farther a supplier is from the consumer, the higher the potential for variability (e.g., oversupply or undersupply). For example, when a person orders a beer in a bar, there is direct communication between the patron and the bartender. The order is placed and then filled. But this immediacy becomes increasingly more difficult moving out along the supply chain. If there is a sudden surge in demand for a type of beer at a bar, the bartender will be instantly aware of this; it will take longer for the brewer of the beer to realize this, and even longer for the grower of the hops (and by the time they react to the changed demand, it may have changed again). In traffic, Carlos Daganzo has pointed out, cars flow through a bottleneck rather smoothly; cars far upstream of the bottleneck, however, experience wide “oscillations” in speed. They are less aware of the actual conditions of supply and demand than those cars moving through the bottleneck. See “The Beer Game and the Bullwhip,” ITS Berkeley Online Magazine, vol. 1, no. 2 (Winter 2005).
by the car following them: Gary A. David and Tait Swenson, “Identification and Simulation of a Common Freeway Accident Mechanism: Collective Responsibility in Freeway Rear-End Collisions,” CTS 06-02. Intelligent Transportation Systems Institute, Center for Transportation Studies, University of Minnesota, April 2006.
Traffic Page 43