Underwater Inspections: 10 Things You Should Know (And Why You Should Care)

Visual inspection intelligence with computer vision and data science

Machine vision, also known as Computer Vision, is the technology and models used to provide image-based and videos analysis intelligent automated inspections.

It is a system that uses visual computing technology to mechanically “see” the activities that take place one by one along the air, sky & seas.

What is underwater visual asset Inspection and why do I need it

Visual inspection intelligence can be land, sky & sea. This article will speak more about underwater inspections for offshore and marine applications.

A platform solution like Ookeani luure by Pagarba Solutions is one example of a AI powered system where unique “built-in” machine vision technology allows users to instantly inspect the underwater environment and carry out inspections audits and certifications without additional manual labor or professional divers.

This technology gives users time to carry out multiple inspections and identify sources of noise and leaks, saving valuable time and money.

Benefits of Underwater Visual Inspection

Underwater visual inspection systems have many benefits : Get visual inspection data for your production; on-the-fly from several cameras (instead of manually viewing the video data at a frame by frame rate) ; Fast and elegant integration, customised to your needs;

A decentralized or Centralized system to interpret and classify an entire product inventory at the same time.

How can underwater visual asset Inspection intelligence help me today ?

Underwater Inspection is a method of real-time visual inspection with the goal of detecting leaks, dangerous objects, irregularities or any hazardous situations. The underwater inspection is done using computer vision. It is used to identify anomalies and failures in parts, vessels, cargo, pipelines or facilities.

This system collects and presents data about the anomaly or failure. The presentation of data makes the information available to the people who are involved in fixing the problem or asset.

A direct benefit is that it improves the inspection process from several hours to a few minutes. The underwater visual inspection can be applied in industries like oil & gas, petrochemical, aerospace and defense, ship building, energy & power and pipelines.

Other Awesome Benefits of Underwater Visual Inspection intelligence

Various 3rd party software and data feed integrations are built to integrate with Pagarba Solutions Ookeani-luure software. No complex integration and process setup. Multi-platform integration. No technical issues. No limit of applications. Remote access via tablet or computer. Far superior method of inspection. No need to move the machine at all. No extra cost. Powerful underwater inspection ROV drone.

Underwater Inspection Process

Because of the nature of work at various depths and currents and weather , operations require constant monitoring for safety and integrity. To achieve this, operators need to place equipment at specified points in the water in order to obtain surface views. Or hire dedicated professional certified divers.

The first and most basic thing to look at is the integrity of the ship. This entails a thorough visual inspection. This part of the inspection is performed by an inspector diving down through the water to the bottom. The inspector must go through all of the same processes of regular operations. So, at any given time, at least one person is down at the bottom and a crew is up above monitoring the situation. And don’t forget about all the regulations and scheduling time and resource commitments. The equipment is needed to ensure that the human does not enter water at too high a pressure or current or wind and risk an environmental hazard or human casualty.

How to Inspect Underwater Assets

1. On the spot inspection with underwater ROV drone : The RoV drone goes under water to inspect the assets, collect data , process data, analyze it, run and train the AI models , return a visually stunning dashboard report that scanned 24 hrs worth of video analysis in under 30 minutes. Even real time if less intense.

As soon as the drone comes back to the surface, it’s time to recharge. There are lots of features that this machine vision technology offers: a. The ability to find even the smallest irregularity b. The ability to see surfaces underwater c. The ability to see shapes of submerged objects d. The ability to control the instrument and environment to get a better result e. The ability to inspect with more accuracy and quicker, compared to human sight f. Faster automated processing. G. Thermal and sonar complex analysis humans are incapable of doing.

Why does machine vision and AI help visual inspections?

The good news is that visual inspection has been around for centuries, first with people and animals , next with Audio and video cameras. It was at first considered too expensive and too complex for visual inspection to become the mainstream. However, as the inspection technology started to be integrated with modern visualization, automation and industrial control systems, AI-based technology in visual inspection and robotics have started to become widely popular. The bad news is that most of the machine vision and AI tools have limitations in the quality of the inspection they can provide. The good news is Pagarba solutions is the visual inspection intelligence experts.

Image recognition basics

Image recognition is a technique for automatically identifying the objects in an image and then classifying those objects into segments , categories and metadata. High definition images and objects are represented using pixel values. Image recognition algorithms can extract pixel values that represent an object or a color and can identify those pixel values.

A common approach is to employ a convolutional neural network to perform image recognition. Convolutional neural networks are a computational structure that learn to map features to representations of the object or the color. Neural networks can also be categorized into convolutional neural networks, net-based convolutional neural networks, and Recurrent Neural Networks amongst many more.

Visual inspection Object detection basics

For smart manufacturing applications, real-time visual inspection is a significant application of image-based machine vision. There is a wide variety of inspection solutions that can be implemented in a smart manufacturing environment. The most common inspection systems utilize the live image of a physical asset to perform a visual inspection. Each single-camera-based inspection system includes components that can: a. adjust object focus; b. remove foreground objects; c. correct for lighting; d. correct for exposure; e. correct for scene shape and contrast. The above are general image-based inspection systems, as defined by ISO/IEC 14443:2008.

how machine vision improves your inspections and maintenance cycles

It reduces downtime and enhances productivity as well as provide an automated step-by-step visual inspection intelligence predictive audit and risk video analysis planning routine did better maintenance scheduling. Help build your inspection and maintenance schedule by uploading inspection workflows, calendars and video search & analytics content discovery and correct defects.

Isolated Visual Inspection for industrial technology that has been integrated into many mobile cameras, replacing the need for an additional standalone inspection and maintenance tool.

Why Pagarba Solutions ?

Thinking machine vision underwater inspection is quite complex in terms of implementation. Although there are technologies that can inspect underwater with decent vision such as: Planar Wavefront Scanning, Direct Imaging Radars , Multi-Spectral Laser Scans, Scanning Electron Microscope , CMOS Cameras , etc. The proprietary machine vision and data science technology built in the Ookeani luure platform by Pagarba solutions is able to penetrate the water, listen to the air and se and not only scan the underwater environment, but build a 3d visual model of the sustainable water ecosystem figure.

Water Scanning systems or multi-spectral systems can be put in the water, from many different angles, to avoid waves. But sometimes, the saltwater can be removed by the water and this also affects the quality of the images. Pagarba Solutions finds innovative ways in solving these problems.


The items examined in this article are ways in which you could inspect underwater visual assets, not the only things you should examine to ensure that your overall asset lifespan is as long as it can be. The findings will help you go about your inspection and product inspection processes to ensure that you get what you want.