If your business relies on consistent quality control, traditionally undertaken by a human inspector, you will may well already have heard about machine vision technology.
If it sounds complicated, that’s because the various technologies that go into creating a good machine vision system can be highly sophisticated. However, the benefits it is bringing to increasing industrial sectors are something every business leader will understand.
By its simplest definition, machine vision is when a machine mimics a human inspector but uses technology that surpasses human sensory capabilities. It is used in all sorts of ways, from quality assurance in factory manufacture to refereeing decisions in competitive sports.
In purely technical terms, machine vision brings together existing technologies such as high-resolution cameras, image processing and big data to enable a computer to interpret visual information. Without any human intervention, these computers can then set in motion other automated processes, such as instructing robotic systems to act.
Machine vision systems are currently widely used in manufacture, or any industry that requires accuracy and reliability in guidance, gauging, inspection or identification processes. Implications for the future of machine vision are enormous, with benefits rippling out across healthcare, agriculture and the wider business world. With improvements and advances happening all the time, machine vision is poised to further transform production processes, drive efficiencies and improve standards in ways we’ve never seen before.
Reasons for businesses to use machine vision
Since its inception in the 1950s, through the first machine vision course at MIT in the 1970s, to its many uses today, machine vision has gone from conceptual computer science to a key feature of manufacturing. The latest systems offer incredibly flexible solutions that help identify defects, sort products and complete a variety of tasks more quickly and efficiently than humans ever could.
Today, any business that relies on producing items to a specific standard could potentially benefit from machine vision as part of its manufacturing process. The creation of any piece of intricate equipment for use in medical devices, for instance, or sensors that will end up in a car or aircraft, require pin-point precision and reliability: there is no room for error.
Heavily regulated industries, such as large pharma, are increasingly relying on machine vision to detect anomalies in their medicinal products. Indeed, new EU Good Manufacturing Practice (GMP) guidelines that are currently being passed, dictate that human inspection will not be accepted once the new legislation is finalised.
With risk minimisation of paramount importance in the pharmaceutical sector, machine vision will therefore undoubtedly play an increasingly central role in removing risks associated with human error.
Is machine vision only suitable for heavy industry?
Machine vision’s most obvious applications do suit businesses that use heavy machinery, or at least businesses with a factory type facility operating a production line.
Typical applications include:
- Automotive construction
- Engine part inspection
- Label inspection on products
- Checking medical and other devices for defects
- Sorting food
- Verifying data codes
- Packaging inspection
- Reading bar codes
- Checking engineering components
It was for heavy industry that machine vision was developed, traditionally using hardware as its inspiration, with the software necessary for its use trailing behind. A good comparison might be the software that came with your scanner: think unintuitive, clunky and with limited functionality.
However, machine vision is increasingly finding its way into other industries. There are apps now available for free that turn a smart phone into a little machine vision system. Take a photo of a flower, and one app uses the internet and software to analyse and identify the flower and then tells you what it is.
This tiny system uses the camera in the phone to capture the image, the phone’s computer to search online and analyse the image, and the phone’s screen to relay what it has learned about the image back to you. Machine vision is being used in autonomous cars, in ticketless car parks that capture your registration number on entry and all sorts of domestic IoT products.
While heavy industry is still the leader when it comes to consistent machine vision application, sophisticated computer technology and equipment developed for the wider market is now improving its efficacy and bringing down costs. This is great news for smaller industrial innovators that need more flexibility and agility in their production lines to satisfy increasing customer demand for customisation.
Machine Vision 2.0
Advanced high pixel-density camera sensors, development of vision algorithms, deep learning and statistical tools plus the proliferation of cloud computing, are just some of the new technologies driving an evolution in machine vision systems.
Empowered by big data style analytics, connectivity and unlimited data storage, machine vision has already come a long way since it could only offer a pass / fail inspection facility. It is no longer simply a reactive tool, it is becoming a data collection tool that supports the prevention of defects.
The data collected by the next generation of machine vision systems is already becoming part of the larger business intelligence network. It will eventually help businesses better understand their complex manufacturing processes and drive more effective collaboration along the entire supply chain – and between engineers, production workers and business management – in real time.
To understand more about machine vision and how it can drive business efficiencies today, visit our dedicated page.