When asked which of the traditional five senses they would most regret not having, human beings generally choose sight. Is the same true for industrial equipment? In this article, John Young of EU Automation looks at some of the latest trends in machine vision in metalworking.
The global machine vision market is worth approximately $9.6 billion and is expected to grow at a compound annual growth rate (CAGR) of 6.1 percent over the next five years, according to research by MarketsandMarkets. In the APAC region, demand is being boosted by manufacturers turning to artificial intelligence (AI), Industry 4.0 and the Industrial Internet of Things (IIOT), all of which benefit from machine vision capabilities.
Definitions of machine vision vary, but most involve the idea of using technology to extract information from images on an automated basis. Machine vision does not refer to a single piece of technology, but rather to multiple technologies, hardware, software and integrated systems.
Deployed in the right way, machine vision can help automate the repetitive and dull tasks traditionally carried out by human workers. For example, sorting parts on a conveyor by colour. Machine vision allows these jobs to be performed at higher speed and with greater consistency, resulting in more efficient quality control, reduced waste and higher yields for manufacturers.
Machine vision technology has been used in manufacturing applications since the 1980s, but there have been barriers to more widespread adoption. Traditionally, perhaps the two key difficulties for manufacturers contemplating adopting machine vision have been cost and the difficulty of installation. As well as being prohibitively expensive in many instances, the equipment often needed a trained and specialized system integrator to set it up.
The latest generation of machine vision technology has gone a long way towards solving these dilemmas by providing systems that are vastly less expensive and much quicker and easier to install. Furthermore, while some machine vision systems might have required hours of ‘training’, which involves feeding images of defective and non-defective parts to the system to allow it to improve its identification capacity, modern technology incorporates machine learning algorithms. This introduces a substantial level of automation into the process.
Another traditional hurdle for the adoption of machine vision in quality control has been the complexity of identifying defects. Take aluminium as an example. Distinguishing between genuine defect and an appropriate level of variation in this alloy is more difficult because of variations in colour and other properties of the material. Many manufacturers would persist with manual inspection, even when inspection errors were made in a quarter of all cases.
Today, machine vision technology is sophisticated enough to make it a commercially viable alternative to human inspection even in more difficult scenarios such as these. Although well-suited to inspection and quality control, modern machine vision systems are multi-purpose and multifunctional. Machine vision can simultaneously offer other benefits like checking OCR codes or monitoring factory equipment as part of a predictive maintenance program.
Automation in metalworking is growing and this growth is strongest in the APAC region. Cobots, or collaborative robots that can work safely alongside human workers, are a good example of this. Cobots are a key area of development in metalworking and are finding new uses in applications like welding, assembly and sorting.
For other exclusive articles, visit www.equipment-news.com.
Check these articles out:
WANT MORE INSIDER NEWS? SUBSCRIBE TO OUR DIGITAL MAGAZINE NOW!