Although most of us take it for granted, driving is a difficult task. Researchers developing automated vehicles are relearning this every day. The driver must constantly control the vehicle, follow the road, predict the motions of other road users, and decide whether an action is needed to avoid them, while also navigating to a destination. There are limits to how much information a driver can process at a time. When these limits are exceeded, mistakes happen.

As drivers gain experience they expect consistency, meaning that things will happen as they have before. For example, most drivers expect that a red light on a traffic signal will be proceeded by a yellow light. Drivers look at an upcoming curve and adjust their speed based on similar looking curves they have driven. This is called expectancy. If a signal changes from green to red without a yellow interval, or a curve radius suddenly gets tighter halfway through the corner, the driver’s expectancy is violated, and the driver may react in an erratic or improper way.

From the Cornell Local Roads Program workbook “Solutions for Safer Roads and Streets,” driving can be modeled as a closed loop, involving information that leads to a decision to perform an action. The results of that action provide more information, starting the process over again.

There are two aspects to the information phase: detecting information and recognizing its meaning. Obviously, there is a delay from when drivers process some type of information that requires a decision, followed by some form of action.

A simple example is a driver approaching a stop sign at an intersection. The information phase is when the driver sees the stop sign. The second phase, the decision phase, is what action will the driver have to take. We know exactly what action this stop sign requires, namely coming to a stop and then observing when he/she may proceed. This is a simple one that all drivers do many times daily.

Fulfilled expectancy leads to quicker and more correct reactions. Violated expectancies increase driver errors and reaction times. It takes more time to figure out what is going on when something is unexpected, and, if the extra time is not available, crashes may occur. That is why expectancy violations cause problems.

An illustration of an unexpected situation where driver expectancy is violated might be observing an oncoming vehicle that is headed toward you in your lane as it passes another vehicle. This information is unexpected, as this should not happen — your expectancy has been violated. In this case, action may be required as fast as possible, but the decision phase may be delayed because of the unexpected situation caused by the oncoming vehicle in your lane. When you first see that vehicle, you will likely experience a “denial” period, hopefully only a fraction of a second but it could easily be more, as you don’t expect to be in this situation. Once you process the information, the decision phase may include multiple possible actions. You could just brake, you may have to take the ditch to avoid a head-on crash, or you might just be able to slow down enough for the oncoming driver to complete the pass.

All this takes time, and time means distance covered by your vehicle. Expectancy plays a big part. Hopefully it helps to understand the three phases of the driving task.