Digital Transformation Is Changing the Face of Visual Inspection

Industry 4.0 is the next revolution, driven by data and characterised by the Internet of things (IoT), cyber/physical systems, and cloud computing.

This digital transformation is proving to be a challenging journey for many organizations as they move to create incremental efficiencies and reduce costs by converting traditionally manual, time-consuming tasks into digital processes—and, in the longer term, seek to harness the potential of computer analytics, such as artificial intelligence, and create meaningful insight from numerous data sources.

As an example of how industry is advancing new technology applications, AllAssets Version 2—software that provides companies with an understanding of risk in relation to plant performance and reliability—was launched recently at the ARC Forum Analyst Conference in Orlando, Florida.

The product is a cloud-first approach with open integration standards to enable users to connect to plant data sources easily and securely with seamless integration. This allows for more-accurate risk mitigation and improved resource use. It houses preconfigured templates, and a model-builder tool enables users to build bespoke models without the need to engage third-party providers, bringing a completely new approach to asset-performance management. Users are able to configure risk models for their own environments in days or even hours and without any software coding, saving time and money. 

In a climate of ever-increasing scrutiny on performance and risk, software developments such as this provide industry with confidence that operational risk can be managed successfully and assurance for operators that their plants can be optimized at a technical and engineering level for better performance against regulatory requirements.

But, while we may think of digital data as predominantly alphanumeric, a more ubiquitous data format is likely to play a significant role in this transformative environment—visual data.

Visual inspection is the oldest and most-natural of techniques for understanding and confirming the nature and condition of any asset and is the primary means for verifying the validity of information obtained through other sources and senses. For that reason alone, it will remain a significant element in any complete inspection solution.

Visual information is also an incredibly powerful tool—static and dynamic images are universal and, unlike language, do not require translation. Its meaning is collective and easily understood.

But, unlike manual inspections, using digital techniques for image analysis such as machine learning can unlock solutions at scale, recognizing changes in asset condition and identifying defects; across multiple assets; and in diverse geographies. Cloud-computing processing and storage power now affords us the speed and reduced cost that was previously unfeasible for a human to achieve.

Harnessing the value of this data can and will change incrementally our approach to inspection, reporting, and analysis. Data of this type means we can conduct inspections remotely; at scale; with repeatable consistency and quality; and, most importantly, securely through cybersecurity systems.

The continuous nature of digital data collection, collation, and analysis presents the ability to monitor and benchmark assets throughout their life, not just as individual assets but also across asset fleets and portfolios. The results of such descriptive analytics not only enable us to see patterns and trends, leading to predictive insights and actions, but also allow us to inform the design phase better, bringing the field back into the laboratory and completing the loop between design and operations.

While it is often assumed that a physical inspection only involves looking at the particular items of interest—and that a picture tells a thousand words—in reality, a human conducting an inspection will be considering a number of factors in their evaluation of status or condition, including sound; smell; touch; and, importantly, context. We use all of these, in addition to movement for perspective and other supporting tools (such as a light or a hammer), to test and validate our assumptions of what we initially see.

As we move to digitize our visual inspections with a variety of image-capture devices, fully understanding the strengths and limitations of the approach is important to move truly from a qualitative to a quantitative assessment with confidence. This approach should be one that incorporates the natural human need for validation and capitalizes on the human skill of evaluating complex data to make reasoned decisions. Only then can we fully capitalize on the possibilities of digital data analytics with confidence. That is why we talk about smart solutions shaped with human intelligence.

While, in reality, the complete digital picture will be built from a combination of different sources and solutions, the adoption and implementation of these will need to be in stages as each is developed, evaluated, and commercialized.


Don't miss out on the latest technology delivered to your email monthly.  Sign up for the Data Science and Digital Engineering newsletter.  If you are not logged in, you will receive a confirmation email that you will need to click on to confirm you want to receive the newsletter.