The real risk of using AI in transportation? Not moving fast enough
April 03, 2025
The cost of waiting on AI in transportation is too high. Here¡¯s why it¡¯s time to act¡ªand how it can deliver better outcomes for our communities.
We¡¯ve spent years working with artificial intelligence (AI), from water to urban design. Because of that familiarity, we know it works.
But the transport sector isn¡¯t moving fast enough to maximise the benefits of AI in transportation. Process, policy, and plain old hesitation are holding us back. In our view, the cost of waiting to implement AI in transport¡ªor any other public infrastructure¡ªis too high.
AI is already reshaping the way we design, deliver, and manage infrastructure on both sides of the Tasman¡ªand around the world. The tools are ready and the impact on our communities is measurable. But we need to get moving.
Safer roads start with smarter maintenance
Australia invests more than $30 billion a year in its road network, with up to 40 percent going toward maintenance. That¡¯s a huge spend¡ªand one that could go much further with AI. In New South Wales, several councils are trialling AI tools to automate road inspections. Instead of sending crews to each site, sensors and cameras capture surface data. AI models then flag defects, suggest repairs, and predict wear and tear before it worsens.
AI-driven transportation could pave the way to a future without traffic jams.
The New Zealand Transport Agency Waka Kotahi (NZTA) is already putting this approach into action. With digital twins, asset owners can monitor roads, bridges, and tunnels remotely¡ªno need for in-person site visits. We¡¯ve seen how effective this is in water networks. For example, the ESRI utility network allows operators to pinpoint pipes, assess their condition, and isolate contamination risks before they become a public health issue. It can work just as well for transport.
We can also use AI to model how people move and pinpoint high-risk zones. In New Zealand¡ªwhich has the fourth highest road deaths per capita in the Organisation for Economic Co-operation and Development (OECD)¡ªdata modelling identified that 80 percent of the country¡¯s roads should have lower speed limits. This is what AI in transportation does best: using data to back smarter, safer recommendations.
Looking ahead, AI in transportation could take us even further. Imagine logging your commute and receiving a personalised departure time¡ªbecause the system has already staggered travel across your area to avoid congestion. Apps like Google Maps are a start, but connecting the full network of AI tools could unlock something even bigger¡ªa future without traffic jams.
Why AI in transportation is a lifeline in disasters
Cyclones, fires, and floods don¡¯t just damage infrastructure¡ªthey sever vital lifelines during times of crisis. A washed-out bridge or flooded arterial isn¡¯t just an inconvenience. In those moments, speed of response matters¡ªand AI helps us move faster with more clarity.
Over 3,000 instances of storm damage highlight the immense impact of Cyclone Gabrielle on the transportation network.
After Cyclone Gabrielle wreaked havoc on New Zealand in 2023 there were more than 3,000 instances of storm damage to the transport network. AI helped us respond quickly. In Hawke¡¯s Bay, we created a geographic information system (GIS) where inspection and assessment data could be logged in real-time, allowing immediate decision-making and reporting.
In the US, our teams have partnered with Tennessee¡¯s Economic and Community Development Department to respond to natural disasters and predict their impact with our Flood Predictor tool. It combines AI with climate and terrain data to forecast where floodwaters will go. That¡¯s now informing everything from emergency planning to major infrastructure decisions.
With climate change making ¡°once-in-100-year¡± disaster events a more regular occurrence, we need to get ahead to make sure we respond quickly before and during emergencies. Using AI in transportation infrastructure means we can do this faster than ever.
AI-powered flood prediction tools are analysing data from weather stations, satellite imagery, and historical patterns to forecast flood risks with remarkable accuracy.
Bringing clients and communities along for the ride
AI is completely changing how we approach projects and engage our clients and communities in the process.
With digital twins and rapid design tools, we can now test dozens of options in the time it used to take to draft one. We can deliver three times as many viable concepts as we could without AI. We can test ideas and show clients those that will have the best outcomes in their communities.
By using 3D rendering AI models, we can take a sketch or a few descriptive words and turn them into a visual in seconds. For someone outside the industry, that makes the abstract real. They can see how something will work in their communities. When people see a design and respond to what they like or don¡¯t, we¡¯re not guessing¡ªwe¡¯re collaborating. And that makes a measurable difference in the outcomes we deliver.
With digital twins and rapid design tools, we can now test dozens of options in the time it used to take to draft one.
From road safety to disaster recovery: Why using AI in transportation can¡¯t wait
The risk isn¡¯t that using AI in transportation will fail. We¡¯ve already seen the results. The real risk is falling behind.
We¡¯ve heard the concerns. What if the tools aren¡¯t perfect? What if the policy isn¡¯t ready? But here¡¯s our take: waiting comes at a greater cost. It wastes public resources and means we¡¯re not moving fast enough to improve the safety and resilience of our communities.
Big tech companies are already circling the infrastructure space. If we don¡¯t lead, we¡¯ll see them taking charge in designing our transport infrastructure, without human Â鶹´«Ã½ and experience behind it. We¡¯re not being ¡°territorial¡± about transportation design¡ªit takes engineers with AI, not AI instead of engineers. We must harness that experience to drive AI adoption so that solutions are grounded in a deep understanding of our built environment. This isn¡¯t about rushing on board out of fear of missing out; it¡¯s about confidently steering the technology where it counts.??
Using AI in transportation shouldn¡¯t be a replacement for the experts. But it can make our work stronger and faster¡ªand deliver a better outcome for everyone.