EnviroNZ trials computer vision for hazard detection

By on
EnviroNZ trials computer vision for hazard detection
EnviroNZ computer vision hazard detection system.

Environmental services firm EnviroNZ has kicked off a 5G-based multi-access edge compute (MEC) pilot for a computer vision system to spot health and safety risks at its resource recover centre.

The computer vision system came about after members of the public tried to retrieve things from one of EnviroNZ's resource recovery centres, despite the area being clearly signposted as no-go with excavators at work, the company's chief executive Chris Aughton said.

Nobody was injured but the incidents led to EnviroNZ contacting Spark subsidiary Qrious, asking the company to develop an artificial intelligence solution to anticipate and monitor breaches of site safety.

Qrious developed a hazard detection system that detects when people get too close to the excavators at resource recovery sites, triggering alerts if they are too near the diggers.

Spark enhanced the Qrious solution with 5G, using the wireless data protocol to transmit the feed to an Amazon Web Services Snowball Edge device with storage and compute capabilities, the first to be deployed in New Zealand, the cloud giant's country manager Tiffany Bloomquist said.

Spark also connected the video camera for the Qrious application with 5G, for better performance, allowing hazards to be identified more quickly.

Using MEC over 5G provides fast throughput and low latency, along with high levels of reliability, Spark's technology director Mark Beder said.

The EnviroNZ is Spark's first MEC project pilot.


Got a news tip for our journalists? Share it with us anonymously here.
© CRN New Zealand

Most Read Articles