AI on the road: cutting emissions, congestion, and chaos

In simple terms, how would you define “smart transportation”?

Dr. José F. Papí

Smart transportation means using methods, technologies, and/or standards that make transport infrastructures safer, more efficient, and more sustainable. On the digital front, “smarting” a transport infrastructure might involve combining technologies like sensors, AI, and data platforms with policy and planning, so for instance roads and other transport networks can adapt in real time to changing conditions, much like a smart thermostat adjusts a home’s temperature. It is not just about gadgets—it’s about coordination and smarter decisions at every stage of the infrastructure lifecycle.

How is AI already reshaping how we travel, and in what domains do you foresee it having the greatest impact on businesses and the general public?

Nowadays, there are already some limited AI applications, such as scheduling tools, automatic camera detection and tracking or ticketing systems. Due to privacy concerns, however, their applications are very limited in the EU for now. Outside of the EU, namely in China, the application for AI is already much more advanced and extensive, including automated auditing of infrastructure items, traffic management, and traffic violation detection and tracking. AI is already streamlining traffic signals, detecting road incidents with cameras, optimizing maintenance schedules, and enabling predictive planning. The biggest impacts will be in traffic management, maintenance, and public transport, where AI reduces delays, lowers emissions, and improves service reliability.

Tell us more about the findings of your technical report on the application of AI on transport infrastructures?

Our latest STA Technical Report highlights AI's value across the full lifecycle of road infrastructure—from planning and design to upgrades. Key technologies like reinforcement learning, federated learning, and digital twins are being used not just for efficiency but to enhance safety and sustainability. The report also highlights the need for strong governance, data protection, and public trust as critical enablers.

Do you have any specific case studies that highlight both the existing value and the challenges of AI in transport?

There are several of them in Europe. For instance, Dortmund (Germany) uses digital twins for predictive traffic management. Porto (Portugal) combines traffic AI with environmental sensing for climate planning. Tallinn (Estonia) is piloting AI-equipped buses to detect illegal parking. These cases show that while AI improves flow, safety, and compliance, challenges include system complexity, sensor reliability, and public acceptance.

What role does AI/smart transportation play in the campaign for a more sustainable world?

AI helps reduce emissions by minimizing congestion, optimizing transport flows, and enabling predictive maintenance—meaning less wear and fewer emergency repairs. Cities like Porto link AI traffic prediction to environmental goals, showing how intelligent mobility systems support broader climate-neutral targets.

How does the STA ensure that advances in AI benefit all communities and not just those in big modern cities?

At STA, we hold both highways and also secondary roads always at the center of our considerations, since roads represent the veins of the EU transport multimodal system – not only for people, but also for freight of goods. By promoting adaptable solutions, open standards, and knowledge-sharing, scalable models can be developed which are more flexible to local contexts for change rather than one-size-fits-all platforms.

Smart transportation will impact and be embedded within critical infrastructure. How do you see it affecting our relationship to personal data security/privacy?

Data from sensors and cameras can be very sensitive, since they can include license plates and even faces. Here, we have to rely on and underscore the GDPR and the EU’s AI Act, stressing privacy-by-design and edge computing to limit data exposure. Security-wise, traffic control systems are now critical infrastructure and must be hardened against cyber threats while also balancing carefully the benefit of increased safety for all road users vs. privacy concerns. Ultimately, this evaluation will be up to policy makers and legislative bodies inside the EU – a long but necessary process.

How do policymakers and technology industry partners work together to introduce AI solutions swiftly and seamlessly? And what are the core duties and responsibilities of those tech partners?

Public Private Partnerships (PPPs) are key. Policymakers define goals and legal frameworks; tech partners bring the tools. Their shared duties include ensuring transparency, respecting privacy laws, training local staff, and delivering scalable, interoperable systems. Clear contracts and governance structures are essential to make this a success.

As we make transport smarter, greener, safer, more efficient and more appealing, how do you address the risk that it does nothing to change people’s core behaviors – i.e., we’re not getting more urban residents to walk rather than drive, we’re simply making their cars better?

AI alone cannot achieve behavior change – a multichannel approach is needed. Cities can pair smart systems with incentives for walking, cycling, and transit. AI can prioritize buses at signals, optimize multimodal travel, and discourage unnecessary driving through dynamic tolling but the goal is to shift behavior, not just to speed up cars.

What are the biggest roadblocks to implementing smart transportation?

Key challenges include legal uncertainty, high upfront costs, technical maintenance (especially sensors), data calibration, skill shortages, and public trust. Cities need interdisciplinary teams, phased rollouts, and citizen engagement to succeed. Funding uncertainty and long ROI timelines can also delay progress.

When it comes to tackling the challenges, what keeps you awake at night?

Cybersecurity risks, poorly explained AI decisions, and uneven access to benefits. A single data breach or badly communicated rollout could derail trust. I also worry about rural areas and secondary road systems being left behind if solutions aren’t made affordable and adaptable.

Paint a picture of the future: in a decade from now, how will AI-powered transport change our daily routines?

Expect smoother commutes, fewer breakdowns, smarter maintenance, and proactive safety. AI will coordinate signals, predict traffic, and adapt transit dynamically. Daily travel will feel more seamless—less waiting, less stress, lower emissions. And most of it will happen quietly in the background, without user inputs or feedback.