Kiwa R&D team is developing computer vision tools that use images streams from various sources, such as drones, portable cameras, surveillance cameras and other image sources that may exist, to help manage the industrial sites.

Remote Inspection Analysis

Many facilities have assets in remote and/or difficult to access locations, preventing operators from inspecting them as often as desired.

The massification and technological advances of remotely operated vehicles, specifically drones, make it possible to inspect assets without the need to put people at risk. Kiwa is developing systems that allow remote inspection, using various video streams sources.

The first analysis is performed automatically by Kiwa’s algorithms. If relevant points are found the experts are alerted and all information regarding the asset is displayed.

The Kiwa system is undergoing tests to identify leaks, asset conditions, such as pressure equipment and piping, and in photovoltaic plants for inspection of solar panels.

Remote Hazard Identification

With the ability to use multiple sources of images, Kiwa inhouse algorithms can detect imminent risks, such as fires, equipment anomalies, unauthorized access to conditioned areas, etc.

The entire video stream and processing process is carried out in soft real-time.