Crane safety systems and artificial intelligence

In some cases the application of artificial intelligence has no clear benefit. But for other more complex tasks the power of AI cannot be ignored.

This is particularly true when looking at crew safety systems used around tower cranes. Hong Kong-based ViAct offers machine- and system-agnostic safety platforms for a series of industries.

What Happened

For the construction sector, the company offers an automated AI system which analyses video in real time to both predict and react to potential danger. The system works by taking video from existing CCTV and other sources and feeding that into on-site.

• Leveraging what is described as error-free progress tracking, using the ViAct video monitoring system across a building site can deliver a 95 per cent reduction in accidents.

• This, with a 70 % reduction in manpower costs.

• Early deployments of the new ViAct feature has resulted in a 63 % reduction in heat-related medical cases and 95 % adherence to hydration and rest schedules.

Key Details

A library of more than 200 computer vision modules helps to interpret what the cameras are relaying. From this ViAct is designed to instantly recognise dangerous situations, including missing safety gear, unauthorised access and proximity to working machines.

• Since 1994 the company has been developing technologies designed to improve safety, productivity and operational efficiency on construction projects around the world.

• Maximum range is about 1,000 metres and multiple devices can be active on a single site.

• If there are more than 16 cranes it is possible to add another radio frequency, plus TT3 is backwards compatible with the original TT1.

Why It Matters

Should a violation or near-miss be flagged, the system sends an alert (audio and-or visual or mobile push notifications) to team leaders within one second. All AI-produced data is saved in the cloud for tracking or auditing.

• Other system features include hook height management in prohibited areas, travelling track management and wind speed detection.

• Sensors pick up signals from RFID tags on the machine to determine and complete inter-crane management, including proximity limits.

What Reports Say

Coverage of the story so far points to:

• Continued reporting by Crane & Transport Briefing as more details emerge