Quality assurance is a critical step in all manufacturing sectors, regardless of industry, tier or geography. The emergence of sophisticated machine vision systems has allowed manufacturers to automate this step, but a major barrier remains – complexity. Harel Boren reports
Quality assurance is mandatory in all industries, tiers and geographies. But existing machine vision solutions are often so complex, that end-customers have no choice but to rely on machine vision experts and systems integrators to commission expensive, slow to deploy and highly tailored solutions. But quality shouldn’t be a luxury.
The emergence of fully autonomous – rather than simply automated – technology is one of the major global upcoming trends in robotics and manufacturing.
Thanks to the newest advances in deep learning, AI and electrooptic components, autonomous technology that truly acts and learns like humans is the ultimate goal in many areas of manufacturing.
The accomplishment of this goal in the next two to three decades is destined to add tremendous value to production, significantly adding to yield, productivity and profitability, with manageable upfront investments.
However, a key barrier that producers are facing, without much success to date, is simplicity, achieved at lower costs and in tighter timeframes.
You can have all the technology you want, but if you have no control over it, you will be forced to rely on an external expert. Currently, this is the situation of machine vision technology for industrial quality inspection.
Typically, a systems integrator will choose all the necessary hardware and software on behalf of the client, and assemble a tailor-made solution.
The process of developing, integrating and training such a solution might take weeks or even months, is highly complex and very expensive. The resulting solution might or might not work for all the changing scenarios it has to inspect.
And each time a problem arises, the manufacturer has no choice but to call the expert back, and the whole process begins all over again.
The manufacturer is held captive to the machine vision solution installed. Shopfloor personnel has long realised that these sophisticated and expensive solutions are actually expensive and sophisticated headaches.
In fact, many feel they’ve been turned into ‘profit centres’ whose difficulties continually feed experts and component manufacturers with more and more work – burning more and more budget.
Yonatan Hyatt and I founded Inspekto with the goal of changing this status quo. We wanted to develop a self-contained product, simple to install, that served the end-customer directly.
To serve the people on the shop floor, autonomous technology must be simple and quick to install, at least as effective as traditional solutions, and – why not? – even fun to set up and use. That’s how we developed the INSPEKTO S70, the first Autonomous Machine Vision product.
Autonomous Machine Vision (AMV) is a new category of machine vision for visual inspection and the INSPEKTO S70, is the only universal industrial inspection product on the market. It was designed to be set up by the employees themselves, without any external assistance, in hours, not weeks. It learns the gold standard of any item by viewing only 20 good samples, as they run by on the manufacturing line – without interfering with the manufacturing process. It’s ready to be connected to any PLC, right out of the box, like your smartphone or laptop would connect to any network.
Our motto is ‘just plug & inspect’, which we’ve also registered as the name of the patented technology that powers the S70.
The expert-free setup process allows QA managers to do everything themselves, without relying on machine vision experts nor systems integrators, and costs are typically 1/10th than previous tailored solutions would have sold for. We fiercely believe that in industry 4.0, quality can’t any longer be a luxury, but a baseline accessible to all, and that technology should be used to democratise quality.
Harel Boren is CEO and co-founder of Inspekto.