Elevating Visual Inspection With AI and AR

AI decision-making resembles human decision-making. This is itself a collaborative process, as an evolving AI requires regular input from a human operator.

A mature AI system can provide extra support to workers for an additional layer of quality assurance. In the case of visual inspection, AI technology may be able to point out features that a quality engineer may have missed. The worker can then act on or ignore these features based on his or her experience and insights. The attention they used to spend recording their inspection results can be redirected to doing the task faster and more efficiently, as the AI ticks all the boxes during the inspection.

Augmented reality uses digital interfaces to allow an operator to interact with a computer system without taking their eyes off the environment. The operator can now look through a screen and see 3D-augmented guidance on the physical world, rather than shifting their attention from the environment to paper instructions.

There are many benefits to implementing this technology in the workplace. It streamlines the operator's task by eliminating the need to constantly divert their attention. It also means they have improved situational awareness and workplace safety because they can look at through a screen without losing sight of their surroundings.

There are opportunities for workforce automation, but visual inspection is not one of them. Whenever a decision needs to be made, it should be made by a person, and the technologies involved don't change that. AI technology offers support and insights, while augmented reality provides in-context visualization. Augmenting rather than automating tasks brings out the best in expert operators by providing the resources they need to do their jobs to the best of their abilities—and beyond.

It’s difficult for an organization to meet critical KPIs as their experts retire and new, inexperienced workers join the company. Onboarding and training can be time-consuming and expensive, especially as machines and processes become more complex. AI-enhanced AR instructions can be authored by the few remaining experts you have, and then can be scaled out across the organization to help new hires. These Digital Mentors provide the same expertise as your most seasoned workers, but in a much more hands-on and efficient way.

The more connected a workplace is, the more it can gain from implementing AI and AR solutions. Having more virtual models and maps available to an AR system makes it more robust. An AI program with access to more statistics and resources will be even more useful to your workforce.

Perhaps more than anything else, augmented reality, working in tandem with other connected devices and programs, will make more information available to the modern worker in more actionable ways than ever before. The way the average frontline worker interacts with their workplace and organization will improve with this freer flow of readily available information in easily accessible formats.

The actual process of visual inspection is the same whether AI is involved or not. In both cases, the inspection engineer walks around a large object, such as an engine block, or rotates a small object, like a spark plug, looking for common points of failure. The inspection engineer follows a prescribed workflow, noting how each unit inspected compares to an ideal model. This information is then shared to optimize the process and address recurring problems.

The difference is that AI technology guides the inspection engineer through the workflow, stating inspection points on each unit, suggesting potential fixes when necessary, and automatically completing and submitting reports along the way.

Both artificial intelligence and augmented reality transform the way inspection engineers can access information. AI helps workers access more information faster than ever before. AR acts as an interface between the worker and the vast amount of information that AI makes available to them. It helps them sort through all that information and view it in a way that is easy to understand and use.

AI reduces inspection cycle time in several ways, depending on the relative familiarity of the product and the relative experience of the inspector. AI helps inspectors become familiar with a new workflow or product by visually guiding them through the process and automating report generation. Even for an experienced inspector, AI can pick up small problems from a distance—including elements that an inspector would need to see up close or even test manually, such as the tightness of a cap or plug.

Most industries, from fabrication and manufacturing to assembly and customer service, can benefit from visual inspection. Visual inspection can be the last step before a unit is shipped, or the first step when a component is opened, but before it is used in assembling a larger system. It can also be used by field service technicians to help customers and clients troubleshoot problems.

You don’t need a lot of data to get started with AI. But the more information you have, the more useful AI will be on a faster scale. The first step of implementing AI is training it for your specific use case. The more data you train AI with at the beginning, the more useful it will be on its first use.

PTC's AI systems learn from human operator feedback. So even if you don't have a lot of data to start with, you can still benefit from using AI from day one.

AI learns over time to improve visual inspection. As a human operator encounters new product defects and applies new solutions, AI can catalog these cases for future reference. This is one of the reasons why AI and human inspectors work best together.