As Storms Driven By Climate Change Batter South Africa’s Roads, AI Takes Up the Watch

Tomas Kellner

Key takeaways

• Blyncsy’s AI will debut in Africa across 5,000 kilometers of roads in South Africa.
• Tight budgets, climate change, and other factors push the shift to proactive maintenance.
• The project is part of Bentley’s expanding global Asset Analytics push.

Cape Town’s roads are about to get a new set of eyes. Bentley Systems just announced that the Western Cape Government’s Department of Infrastructure will deploy its AI-powered Blyncsy platform across roughly 5,000 kilometers (about 3,100 miles) of provincial roadway. The system will use crowdsourced dashcam imagery and machine-learning models to automatically flag damaged guardrails, missing street signs, faulty streetlights, vegetation creeping into sight lines, and debris—a growing concern as the province absorbs more frequent and severe storms.

The South African project marks Bentley’s first rollout in Africa and its latest step in a global push that already spans the United States and Europe.

The news comes as Western Cape officials confront a familiar global challenge: tight budgets, aging assets, and weather that no longer behaves the way it used to. Recent flooding has cut off entire towns, and the Department of Infrastructure is betting that better data—delivered faster and cheaper than traditional windshield surveys—will help it spend its R4.56 billion transport budget where it matters most.

“Providing safe and resilient infrastructure is the foundation of economic opportunity in the Western Cape, particularly as we manage the impacts of climate change on our road network,” said Johannes Neethling, the province’s chief engineer for transport infrastructure systems. “By integrating Blyncsy’s AI technology, we are gaining a level of visibility that was previously impossible.”

Blyncsy, which Bentley acquired in 2023, has been growing in the U.S. Its imagery map now covers designated roads in every state, with a sample reflectivity score for each capital city. The technology also has been steadily expanding into European markets. The platform was conceived more than a decade ago by Bentley executive Mark Pittman after he found himself stuck in his car behind a broken traffic signal, wondering why agencies still relied on guesswork and ad-hoc inspections to manage a U.S. road network that carries roughly 70% of the goods Americans consume.

Earlier this year, the Hawaiʻi Department of Transportation launched its “Eyes on the Road” program, distributing 1,000 high-definition dashcams to residents whose anonymized footage feeds the same AI platform now coming to South Africa. Hawaiʻi is a useful case study: sun, salty air, torrential rain, lush vegetation, and even volcanic activity grind away at roads that thread across each island, where a single washout can sever a community. As of early January, 898 drivers on Oʻahu had signed up to share footage, with the state actively recruiting more on the islands of Hawaiʻi Island, Maui, and Kauaʻi.

Alabama is also using the technology. The state’s Department of Transportation, which oversees roughly 11,000 miles (about 17,700 kilometers) of state roads and highways, is among the first in the U.S. to adopt performance-based budgeting for road maintenance. The state is now feeding that system with AI-generated condition data, allocating dollars based on what the roads actually need rather than what each district requested the year before. A pilot demonstrated 97% accuracy in identifying more than 50 types of roadway assets and defects, from cracked guardrails to faded pavement markings.

The shift to technology-based systems is being accelerated in the U.S. by a Federal Highway Administration deadline this September that will require every state and local transportation agency to adopt a method for maintaining minimum pavement marking “retroreflectivity.” That’s a measure of how well lane stripes bounce headlights back at drivers, and it’s a factor in roadway departure crashes that rank among the leading causes of U.S. traffic fatalities. Bentley says the Blyncsy approach delivers comparable insights to legacy 360-degree imagery surveys—work that once ran up to $300 per mile and took five months—at roughly half the cost and 98% faster.

The Western Cape launch also extends a busy stretch for Bentley’s Asset Analytics business, which in December closed acquisitions of Talon Aerolytics and the technology and team of Pointivo to expand into telecom towers and electric grids. “They also add to our technical and business momentum and help infrastructure owners and operators improve the performance and resilience of their assets,” James Lee, Bentley’s chief operating officer, wrote of the deals.

“The expansion of Blyncsy into the Western Cape of South Africa represents a pivotal step in our mission to provide global transportation agencies with real-time visibility into the state of their infrastructure,” said Pittman, who is now senior director of transportation AI at Bentley. “Our goal remains clear: to replace historical precedent with AI-driven insights that reduce risk, lower costs, and ultimately save lives.”

FAQ:

Why is the Western Cape turning to AI to monitor its road network?

With climate change driving more frequent and severe storms, the province is deploying Blyncsy’s AI platform across 5,000 kilometers of roadway to automatically flag hazards like damaged guardrails and debris faster and cheaper than traditional manual surveys.

How exactly does the Blyncsy platform detect hazards on the road?

The system uses a combination of crowdsourced high-definition dashcam imagery and advanced machine-learning models. This data feeds the AI, which automatically scans the footage to identify safety issues ranging from faulty streetlights and missing signs to vegetation creeping into driver sightlines.

Has this AI technology been tested successfully anywhere else?

Yes, Blyncsy is already operational across the United States and is expanding in Europe. For example, Hawaiʻi uses the platform via 1,000 dashcams distributed to residents to monitor storm and volcanic damage, while Alabama uses the AI-generated condition data to drive its performance-based maintenance budget with 97% accuracy.