Phys. Rev. Lett. (in the press); preprint at http://arxiv.org/abs/1205.6486 (2012)

Credit: © US NARA

In August 2010, a Chinese traffic jam spanning over 100 km lasted more than 12 days. Research into how such a state might be avoided has highlighted similarities with other jammed systems. Now, simulations by Astrid de Wijn and colleagues have shown that over-cautious driving — motivated, for example, by extreme weather conditions — may induce a dynamical transition resembling the onset of glassiness.

The fact that traffic jams generally have a typical size and finite correlations has hampered attempts to define criticality precisely in these systems. But kinetically constrained models for glassy systems are readily applicable to the problem, in that the dependence of each car's acceleration on that of the one in front provides a natural kinetic constraint. De Wijn et al. found that as density increased, cars decelerated according to this constraint, creating jams and lowering the average velocity of the system.

The dynamic susceptibility, indicating the size of regions of correlated mobility, diverged in their simulations, together with the correlation timescale — inducing the sort of persistence that leads to record 12-day jams. The transition occurred in the limit associated with over-zealous braking, and led to coexistence of free-flowing and jammed traffic, in analogy with a thermodynamical critical point terminating a two-phase regime.