The delay inside
a traffic wave.

A queue can form on a road with no junction, accident or narrowing lane. The vehicles keep moving forward while the disturbance passes backwards through them. Explaining that pattern means following how each response changes the conditions faced by the next driver.

My public preprint, Mathematical Modelling of Oscillatory Dynamics in Circular Traffic Systems, examines this problem through delayed vehicle-following equations. The useful connection is between a local response and the behaviour of the complete loop.

Remove the bottleneck and watch the interaction

In the experiment reported by Sugiyama and colleagues in 2008, 22 vehicles travelled around a circular road with a circumference of 230 metres. Drivers began at approximately equal spacing and were asked to aim for about 30 kilometres per hour while following safely. Fluctuations developed, several cars eventually stopped, and a jam travelled against the direction of the traffic.

The published experiment isolates an important possibility: interactions among vehicles can sustain a queue without a fixed obstruction. It does not identify every mechanism behind congestion on an open road. A ring deliberately removes entries, exits and many of the choices that complicate everyday traffic.

Put the response history into the model

The preprint begins with uniform speed and spacing, then examines disturbances using delay equations and Fourier modes. A mode describes a spatial pattern around the ring; analysing its time evolution asks whether that pattern shrinks or grows.

A delayed response uses an earlier observation of the gap or relative motion. Two vehicles with the same present spacing can therefore respond differently because their recent histories differ. Response strength, spacing and the timing of information become separate parts of the explanation.

This is analytical research. Section 11.3 sets out possible future simulation verification; it does not report a completed simulation or a new road experiment. Its model conditions need to remain visible when interpreting the proposed stability analysis.

Test what a different response can change

A separate experimental programme supplies an instructive comparison. Stern and colleagues studied circular traffic with more than 20 vehicles and showed that controlling the velocity of one vehicle could dampen stop-and-go waves. Their paper, published in Transportation Research Part C in 2018, compares speed, braking and fuel-use measures.

That intervention provides evidence about those controllers under the tested conditions. It does not establish that every automated vehicle will smooth traffic, or validate the particular equations in my later preprint. The link is a research question: which response rules reduce disturbances as they pass from vehicle to vehicle?

Make the comparison observable

A useful evaluation would declare the response rule, the observation delay and the traffic conditions before comparing outcomes. It would follow the disturbance across successive vehicles, alongside journey speed, braking and spacing. An average speed alone can conceal repeated sharp changes within the same journey.

It would also distinguish an initially growing disturbance from a persistent wave. The first concerns what happens close to a reference state; the second requires examining the subsequent motion. A credible comparison needs both a clearly specified model and observations that can challenge it.

This is the broader systems question raised by the traffic loop. What information does each participant act on, how old is it, and what does that action transmit to everyone else? Keeping those links explicit makes the route from an individual correction to collective instability open to investigation.