From Reactive Repairs to Predictive Safety: How AI Is Transforming Road Infrastructure

Mark Pittman, senior director, transportation AI, Bentley Systems, shares how AI-powered digital inventories are shifting transport agencies from reactive maintenance to proactive infrastructure management, improving road safety, climate resilience and operational efficiency across increasingly complex and weather-exposed networks.

When infrastructure failures like collapsed roads, obscured signage or flooded corridors directly impact lives, how does a real-time digital inventory shift the ability of agencies to prevent those incidents rather than simply respond to them?

Mark Pittman: A near real-time digital inventory, powered by technologies such as Blyncsy, enables transportation agencies to move from reactive maintenance to proactive risk management. Using AI-driven computer vision to continuously monitor road networks, agencies can quickly identify issues such as damaged guardrails, faulty streetlights and missing signage.

This continuous visibility allows maintenance teams to address vulnerabilities - including blocked drainage and encroaching vegetation - before they escalate into serious incidents such as flooding, road closures or infrastructure failure.

Pittman: In regions like the Western Cape, where extreme weather is increasingly disrupting mobility, what are the real-world consequences when AI systems can flag hazards before they escalate into road closures or accidents?

In climate-exposed regions, AI-powered hazard detection helps preserve both public safety and economic resilience. Recent flooding events in South Africa have shown how quickly communities can become isolated when transport infrastructure fails.

By identifying early signs of deterioration before severe weather strikes, agencies can take preventative action to reduce road closures, maintain access to essential services and keep supply chains moving. After an event, having an up-to-date record of infrastructure assets also accelerates recovery by helping authorities identify what has been damaged and requires replacement.

Historically, delayed inspections and incomplete asset data have contributed to avoidable road safety risks. What changes when those blind spots are removed through continuous monitoring?

Pittman: Removing data blind spots through continuous monitoring delivers both operational and public safety benefits. Agencies can maintain a highly accurate digital inventory and deploy maintenance resources where they are most needed, rather than relying on periodic manual inspections.

It also shifts decision-making from historical assumptions to near real-time evidence, enabling agencies to identify and address risks before they contribute to accidents or service disruptions.

When maintenance budgets are stretched, how does automated roadway intelligence change what gets fixed first?

Pittman: Automated roadway intelligence allows agencies to prioritise maintenance based on objective risk data rather than visibility or perception. Without AI, limited resources often flow towards the most visible locations, potentially leaving high-risk defects elsewhere unaddressed.

By identifying and ranking critical safety issues across the network, AI helps agencies allocate funding more effectively, reduce emergency repair costs and improve safety outcomes for all communities.

What is the risk of not adopting these systems at scale, particularly in fast-growing or climate-exposed regions?

Failing to modernise infrastructure monitoring creates a growing strategic vulnerability. As populations expand and extreme weather events become more frequent, traditional inspection methods struggle to keep pace with maintenance demands.

The result is a widening backlog of repairs, increased infrastructure failures and a greater likelihood of public safety incidents. Over time, these disruptions can have significant social and economic consequences.

Beyond efficiency, where do you see the greatest direct human safety impact from this technology?

Pittman: The greatest safety impact lies in accident prevention. By using AI and crowdsourced imagery to identify hazards such as damaged guardrails, obscured signs and poor sightlines, agencies can address risks before road users encounter them.

The technology also improves worker safety by reducing the need for manual roadside inspections, limiting exposure to live traffic. In addition, maintaining clear signage, functioning streetlights and visible crossings helps protect vulnerable road users, including pedestrians and cyclists.

While emergency response and long-term infrastructure reliability also benefit, preventing accidents before they occur delivers the most immediate and meaningful safety outcome.

How will lessons from the Western Cape deployment inform adoption in other emerging markets, and how might AI-powered digital inventories evolve over the next decade?

Pittman: The Western Cape deployment demonstrates how advanced data analytics can help governments strengthen climate resilience while maximising constrained budgets. It provides a practical model for other emerging markets facing similar infrastructure and environmental challenges.

Over the next 5-10 years, AI-powered digital inventories are likely to become a core component of smart city strategies. As they integrate with connected vehicles, sensors and broader urban systems, they will support more responsive infrastructure management, helping cities improve mobility, resilience and public safety at scale.