The fast lane, the slow lane
At a glance
Transport networks are often designed around how roads perform, yet real safety outcomes depend on how people behave behind the wheel. We now have the ability to observe driving patterns at scale, helping us understand how culture influences the way drivers respond to limits, hazards and changing conditions. With many countries seeing a rise in serious incidents, this shift matters. GHD’s Sarah Dods and Frank Penry share how their work with connected vehicle data is giving us clearer insight into those patterns and helping guide decisions that support safer journeys.
Over many years, transport networks have been planned by measuring the condition and performance of roads. Newer technologies like connected vehicle data are allowing transport planners to observe how people actually drive and how that behaviour shapes safety outcomes. GHD’s partnership with Compass IoT provides consistent observations of braking, accelerating and swerving patterns, offering a more complete view of how drivers behave in real conditions.
Drawing on their experiences across Australia and the United States, along with insights gathered at the 2026 Transportation Research Board (TRB) annual meeting, Sarah Dods and Frank Penry share how cultural norms are shaping the way we drive and why behaviour data is becoming a central part of transport planning.
Two cultures, two speeds
During his recent work across Australia and New Zealand, Frank noticed a striking contrast between local driving behaviour and what he’s seen in the United States. Australian drivers tend to treat the posted speed as the upper limit. When limits change through work zones or in poor weather, most drivers reduce speed together and move at a similar pace.
In the United States, the posted speed often acts as a lower boundary. Drivers expect the flow of traffic to set the real speed, and slower vehicles are encouraged to move aside. This creates a wider range of speeds on the same stretch of road.
Each country interprets the same rule quite differently. And while both countries have the same aim — safer journeys — they start from very different behavioural baselines that will shape the effectiveness of any intervention.
From incidents to patterns
Crash data tells us what happens after an incident, while behaviour data helps us understand the conditions that lead to risk. When we combine hard braking, swerving, high g-force events and speed variation with crash histories, we can form a clearer picture of where and why risk is building. A layered approach helps us see whether a location experienced a rare event or whether repeated patterns make another crash likely.
Some locations show many near misses, which are early warning signs. For example, we can map dozens of swerving events near a site that’s had only one recorded crash, suggesting behaviour patterns may need attention before another incident occurs. We can also track how far drivers travel before and after a risky event, which helps us understand whether certain parts of a journey contribute to unsafe behaviour.
Behaviour data also shows whether safety treatments create sustained change. Some interventions work at first but lose impact as drivers become familiar with the environment. Connected vehicle insights help us understand whether behaviour shifts hold or regress, especially when drivers become comfortable again. That level of detail gives planners clearer guidance when prioritising upgrades or selecting treatments.
Nudging safer journeys
Behaviour change depends on culture. In Australia, signs that highlight known high‑risk crash locations often prompt drivers to slow down and focus on the road ahead. In the United States, hazard messages may have weaker effects, so enforcement may play a supplementary role. Connected vehicle data helps us compare these responses and learn which messages create consistent improvements in safety and behaviour change.
We can also test the impact of messaging. We have learned that direct instructions to slow down often have limited effect.
Telling drivers to slow down doesn’t work. Telling drivers that there is an accident ahead might be more effective.”
Research supports this. A 2016 paper shows that drivers were more likely to divert routes when the sign displayed ‘accident’ when compared with generic messages like ‘roadworks’ or ‘congestion ahead.’ It certainly creates an opportunity to refine how information is delivered so that drivers can form safer habits.
As autonomous vehicles become more common, we face harder questions about how they should behave around human drivers. For example, should an autonomous vehicle follow the posted limit or match the flow of traffic when those speeds differ? And how should it respond when nearby drivers drift from the expected pattern because they’re distracted or reacting to conditions around them?
While autonomous vehicles follow posted speed limits, they may not align with the human driving patterns around them. The gap raises ethical and engineering questions about how autonomous systems should behave when the ‘safe’ choice conflicts with the ‘legal’ choice on a busy road.
Behaviour data will help us work through these questions with evidence rather than assumptions, and support closer collaboration between behavioural specialists and engineers to make sure interventions reflect real driving behaviour.
Key takeaways and next steps
- Behaviour data helps us see how drivers respond to real conditions and where risk starts to build. It gives us a clearer view of patterns that are not visible through crash data alone.
- Connected vehicle insights help us track whether a safety intervention is changing behaviour and whether that change continues once drivers are familiar with the environment.
- We can use evidence to guide planning and policy work so treatments align with local driving habits and support safer outcomes across regions.
If you’d like to discuss how behaviour insights could inform your safety planning, get in touch to see how we’re implementing this data across projects.