From air bags to zippers, manufacturing defects are costly and damaging to business reputation.
Manufacturers have long turned to inspectors and image recognition systems to spot inconsistencies in products before they’re shipped - with mixed results. If the quality control status quo were fail-safe, recalls would be a thing of the past.
The consequences of failing to identify manufacturing defects
Here’s the scenario: you’re a steel manufacturer with shipments going to customers around the world. In many cases, the integrity of buildings, bridges and other critical infrastructure hinges on the quality of your products.
If there’s a defect in what you’re producing, the best case scenario is that you catch it early, scrap the defective units and find out how to correct the problem when you remake the faulty batch, which is expensive and time-consuming.
The worst case scenario—after the defects have been shipped and installed—is too horrifying to contemplate.
In either case, your business (and possibly other people) has suffered because you didn’t catch the problem until it was too late.
Using Artificial Intelligence to transform quality control
So, how do you avoid this negative experience? Fortunately, we live in a digital era in which AI is becoming more intelligent by the microsecond.
Fujitsu offers an innovative solution called F|AIR that takes image recognition to the next level by integrating with and using AI to:
- Automate at scale for defect detection
- Provide a consistent “expert” assessment of inspection data without bias
- Integrate into production processes to enable remedial action
This enables real-time product quality analysis to detect defects during the manufacturing process itself – not after the fact. By addressing problems at their source, manufacturers can keep production lines moving, workers working and customers happy.
In fact, I suspect having some form of AI safeguard in place to ensure quality control will become standard practice within the next decade: today’s innovation is tomorrow’s standard. Those that don’t adapt to this new reality are likely to lose ground and ultimately shutter their operations.
AI and technology solutions to help solve the skills and training challenge
Exacerbating the issue is the fact that there aren’t enough people available with the experience companies need, and the younger generations are increasingly rejecting the jobs their parents valued and which gave them that experience.
This is sometimes expressed as a problem, but is it?
Visual quality control, for example, is not necessarily the most fulfilling way to spend your time. If we can give the next generation the opportunity to work on more stimulating tasks, using enhanced skills – what Fujitsu calls human-centric optimization – then that’s surely a positive direction.
No surprise, then, that there is an accelerating trend towards “Human to Machine” adoption in manufacturing, which captures the experience of an aging workforce and embeds those skills in a form that is more acceptable to tomorrow’s employees.
Artificial Intelligence is one way in which this is already happening.
For manufacturers, Fujitsu offers next-generation quality control with Fujitsu Advanced Image Recognition (F|AIR), which can achieve savings of greater than 80 percent. With this AI-driven quality control solution from Fujitsu, it takes Siemens Gamesa only a quarter of the time previously required to inspect its wind turbine blades for manufacturing defects, while removing the risk of human error.
Imagine what it could do for your organization.