Traffic jams generally do not require a trigger such as a traffic incident or road activity. Japanese researchers have shown that on a circular track, drivers' small braking differences translate into stop-and-go waves with no obstacles. These ghost jams spread back in waves as each vehicle moves forward. By modeling traffic as a fluid, the math reveals that adding vehicles beyond the optimal density reduces the flow. The same math shows that constantly changing lanes causes disruptions to neighboring drivers, negatively impacting overall traffic flow.
Applied mathematicians use tools such as signal processing, stochastic modeling, machine learning, and control theory to avoid congestion. In a 2018 experiment, smooth acceleration and braking of a single autonomous vehicle reduced fuel consumption by dampening the stop-start wave in the entire circuit. This shows that coordinated decisions can create big impacts with small interventions. Mathematics provides drivers with evidence-based recommendations such as maintaining a safe following distance, accelerating/braking properly, and avoiding lane changes.