Johor Bahru smart city uses AI to manage traffic, potholes and cleanliness

Johor Bahru’s AI‑driven control centre is a concrete example of how mid‑size cities, not just megacities, are starting to operationalise AI as infrastructure. Instead of abstract “smart city” marketing, this deployment focuses on everyday pain points: traffic congestion, illegal dumping, potholes and crime. By wiring 555 cameras into an AI stack that can flag anomalies and dispatch crews, the city is effectively building a basic city‑scale agent system with perception, prioritisation and partial actuation through human teams. ([straitstimes.com](https://www.straitstimes.com/asia/se-asia/in-jb-malaysias-first-leading-smart-city-ai-is-used-to-ease-jams-and-detect-potholes/))

For the AGI conversation, this sort of system is important because it normalises AI as a decision‑support layer for public services. As capabilities improve, the temptation will be to give such systems more autonomy—dynamic road pricing, automated fines, predictive policing—raising governance questions that look a lot like “narrow AGI for cities.” Countries in Southeast Asia are positioning themselves as testbeds for these deployments, which will generate valuable operational data and lessons for richer jurisdictions. It’s a reminder that not all meaningful AI progress happens in Silicon Valley labs; some of it happens when a municipal engineer realises that deep‑learning pothole detectors can cut repair times from weeks to 24 hours.