Elevated hazards are detected not by the naked eye but proactively identified by OpenMV using AI

Falling object incidents are often caused by small cracks and ageing issues that have been neglected over a long period. Manual inspections are time-consuming, high-risk, and unable to detect changes in a timely manner. OpenMV proactively identifies façade hazards through continuous visual monitoring.

In densely built-up urban environments featuring high-rise buildings, ageing façades, expanding cracks, or local spalling—once undetected in a timely manner—can easily lead to safety incidents such as falling objects from height, posing serious risks to personnel and property.

However, traditional façade inspections primarily rely on manual visual checks or high-altitude operations, resulting in long inspection cycles, low coverage frequency, high operational risks, and inconsistent subjective judgements—making it difficult to meet the requirements of continuous safety supervision. By introducing SingTown’s OpenMV intelligent cameras, façade safety inspections can be upgraded from “intermittent manual inspections” to “continuous online visual monitoring”.

The camera can be installed at fixed monitoring points around buildings or mounted on lifting devices or inspection equipment to capture images of critical areas of the building façade at predefined intervals. OpenMV runs edge detection, texture variation analysis, and crack feature extraction algorithms on the camera’s edge device to analyse subtle structural features on wall surfaces and compare them against historical images to identify abnormal changes such as crack propagation, surface spalling, and local damage.

Leveraging the continuous comparison capability of edge AI, the system can not only detect whether an issue exists but also determine whether it is progressively deteriorating. All identification processes are performed entirely on the OpenMV edge device, with only critical results—such as the location of anomalies and trend information—being output. This significantly reduces the transmission requirements for raw image and video data, thereby alleviating bandwidth and storage demands on the system and providing property management or operations and maintenance units with clear, actionable maintenance decision-making support.

Leveraging OpenMV’s on-device AI and edge deployment advantages, safety management of high-rise building façades has shifted from “post-incident investigation” to “proactive early warning”, effectively enhancing the safety level of urban operations.