Transportation Conflicts and AI

In a recent opinion article in Traffic Technology Today, Dr. Nick Reed – founder and CEO of Reed Mobility – discusses some of the potential conflicts between road authorities and individual drivers that can easily occur when AI enters the mix.

Traffic managers and individual drivers alike are becoming increasingly accustomed to advanced traffic management tools, which are supported by AI. But what happens when there are outcome and operational conflicts?

First off: No one wants traffic congestion. Delays, inefficiencies, frustration, cost, missed appointments, obstructed emergency services. These are just some of the negatives when demand for road space exceeds supply. Consequently, there are strong incentives to ‘solve’ congestion.

But it’s not a simple problem to solve. Causes of congestion may be simple, such as roadworks blocking a lane, or complex – emerging from a mixture of demand, environmental, infrastructure, and human factors that contribute to flow breakdowns.

Fortunately, recent years have seen the emergence of AI technologies that are able to take varied inputs and determine optimal outputs in response. So, the capabilities of AI are increasingly being recruited to tackle congestion. It’s a seemingly perfect fit.

Traffic signals can be dynamically adjusted to respond to flow, predictive models can anticipate traffic surges and incidents, while data from sensors, cameras and connected vehicles can be being fed into powerful AI algorithms tasked with squeezing every drop of efficiency from road networks. The goal for road authorities is clear: optimize the system for the collective good.

Simultaneously, individual drivers also have support from AI assistants. For example, navigation apps can optimize routes based on road layouts and live traffic data. These tools are popular for individuals and fleets because they help users maintain progress, avoid delays, and maximize journey reliability.

However, the goals of the road authorities and individual may not always align. A public authority might manage traffic across a city or region, striving for fairness, air quality, and reduced delays at a system level. But an individual driver will follow recommendations made by their device, optimizing only for its single user, which may conflict with societal objectives.

From a human factors perspective, drivers can over-trust navigation guidance. Even when a recommendation seems dubious, they may follow it regardless. Conversely, authorities may struggle to convince the public of the value of routes chosen for the collective good.

What if a more collaborative model could be introduced? For instance, what if traffic authorities and navigation providers could share objectives and data?

These are interesting questions to ask!

Ultimately, traffic management in the age of AI is about aligning digital intentions with real-world impacts. Read the full article in Traffic Technology Today.