Nairobi’s traffic police may soon give way to AI

On Nairobi’s roads, traffic police officers have long acted like human operating systems—stepping into chaotic intersections, overriding traffic lights, waving matatus forward, stopping impatient motorists, and manually holding a city perpetually on the brink of gridlock together.

The government now wants machines to take over. Treasury documents tabled in parliament show Kenya has allocated KES 1.18 billion ($9.1 million) next financial year to expand Nairobi’s Intelligent Transport System (ITS) Phase III, an AI-powered network of smart traffic lights, surveillance cameras, and road sensors that could gradually reduce the need for traffic police officers at major junctions across the capital.

The investment is a sharp increase from the current KES 116 million ($898,180) allocation and signals the government’s growing confidence that cameras and algorithms can better manage Nairobi’s roads than traffic police officers stationed at major intersections.

“The third phase of ITS marks the full integration of Nairobi’s traffic ecosystem,” Kenya Urban Roads Authority (Kura) said in project documents. “It will encompass 125 intersections, linking them to the central control system at Cabanas.”

The system, being rolled out by the Kura, will use artificial intelligence (AI) to monitor traffic in real time, detect violations, and automatically adjust traffic lights depending on congestion levels at specific junctions.

The network will have a command hub at City Cabanas on Mombasa Road, where engineers will watch live feeds from intersections across Nairobi and remotely coordinate traffic flow. The project’s implications could be huge for the city’s five million inhabitants.

For decades, Nairobi’s traffic system has depended heavily on human intervention. Officers manually direct vehicles at busy junctions, especially during rush hour or when traffic lights fail. Now, automated systems will take over much of that work.

Economic cost of traffic

Every smart junction under the project will be equipped with cameras capable of recognising number plates, identifying red-light offences, monitoring speeding violations, and detecting helmet compliance among boda boda riders. The system will also analyse vehicle movement, passenger numbers, and turning patterns before automatically relaying offences for enforcement.

Officials say the project is necessary because the city’s congestion problem has become economically unsustainable. Government estimates released in 2024 showed traffic jams cost Kenya roughly KES 120 billion ($929.1 million) annually through lost productivity and fuel wastage. Commuters spend nearly an hour on journeys that should take minutes under normal traffic conditions.

The government believes AI-controlled traffic systems can reduce those losses by making Nairobi’s roads more responsive and coordinated.

Instead of relying on fixed traffic-light schedules or officers manually adjusting traffic flow, the AI system will continuously analyse congestion levels and dynamically adjust signal timings.

“You don’t have to walk into a junction to adjust signal timings any more,” Kura said in project documents. “Everything happens from the control room.”

The technology mirrors traffic-management systems already used in Chinese cities and European capitals like London, where sensors and AI-controlled intersections are increasingly replacing manual traffic coordination.

A cabinet document approving the project in 2024 described the ITS rollout as a way of “eliminating human interfaces in traffic control,” unusually direct language that hinted at the government’s longer-term ambition.

The project will integrate 125 intersections into the central traffic-control system. Treasury projections show the government plans to spend at least Sh5.3 billion on the programme over the next three financial years, with much of the financing coming from a $185 million concessional loan signed between Treasury Cabinet Secretary John Mbadi and the Export–Import Bank of China in 2025.

Among the intersections targeted are some of Nairobi’s worst choke points, including Moi Avenue and Kenyatta Avenue, Koinange Street and Kenyatta Avenue, Raila Odinga Way and Lang’ata Road, and Limuru Road and Muthaiga Road.

The government says the system will also improve emergency response by instantly detecting abnormal traffic patterns caused by accidents or road disruptions and alerting police and rescue teams in real time.

“If there is an accident, the system detects a change in traffic flow and immediately alerts the control centre,” Kura said. “Police and emergency teams can then be dispatched instantly.”

However, the rollout raises questions beyond congestion. A citywide network capable of recognising number plates, monitoring traffic behaviour, and digitally issuing penalties creates a far more expansive surveillance infrastructure than Kenya has previously operated at this scale.