Machine vision technologies are more important than ever for automation
The need for automation and accompanying technologies such as machine vision and robotics will grow as manufacturing changes in the wake of COVID-19.
Across nearly every industry imaginable, automation technologies help companies keep up with customer demands while improving efficiency, increasing throughput, and hopefully, driving revenue. Despite recent market figure declines from the COVID-19 pandemic and its aftermath, the need for automation and accompanying technologies such as machine vision and robotics will only continue to grow.
Last fall, the World Economic Forum released The Future of Jobs 2020 report, which indicated that by 2025, automation and a new division of labor between humans and machines will disrupt 85 million jobs globally in medium and large businesses across 15 industries and 26 economies. More than 80% of business executives plan to accelerate the digitization of work processes and deploy new technologies, while 50% of employers expect to accelerate the automation of some roles in their companies.
And while it would seem as if this means that jobs are being eliminated, the report also shows that the “robot revolution” will create 97 million new jobs, adding that “communities most at risk from disruption will need support from businesses and governments.” These jobs are expected to emerge in industries that rely on artificial intelligence and content creation—and basically anywhere humans retain advantages over robots for tasks such as managing, advising, decision making, reasoning, communication, and interacting.
The eyes of the system
Several technologies comprise the overarching category of automation, but machine vision and imaging technologies have seen significant technological advancements and increased adoption over just the past few years. Examples include:
- Tasking multispectral cameras with identifying bruises in fruit that are not visible to the human eye
- Using hyperspectral cameras to determine the fat content in avocados
- Deploying the cutting edge in 3D for bin picking applications
- Employing time of flight (ToF) to count and dimension boxes
- Finding the right “eyes” for collaborative robots.
These are just a few recent examples, of course. There has been significant headway made in several vision and imaging areas, beyond just nonvisible and 3D imaging technologies. These would include high-speed imaging, lenses/optics/lighting, embedded vision, line scan imaging, and advancements in image sensors themselves, with notable improvements in areas such as resolution, pixel size, data rates, dynamic range, noise, and so on.
As these technologies progress, so too do the capabilities of the machine vision systems in which they are integrated. For proof, have a quick look at the agenda for Vision Week 2021 compared to The Vision Show 2012. While 2012 featured new products that included CMOS line scan cameras, frame grabbers, 3D scanners, fanless PCs, and interfaces such as FireWire and USB, 2021 offers a look into the latest in deploying vision on the edge, deep learning, hyperspectral imaging, and “4D” vision. While one show was of course in person and the other held virtually, you get the picture.
Upward into the future
Recent A3 figures show, robot orders were up 20% in the first quarter of 2021 compared to 2020, with substantial increases in purchases coming from metals (+86%); life sciences, pharmaceuticals, and biomed (+72%); food and consumer goods (+32%); and other nonautomotive industries (+12%). The first quarter of 2021 was the second-best quarter on record for nonautomotive orders, behind only Q4 of 2020. These numbers seem significant for a few reasons, the first of which is that the automotive industry has obviously relied heavily upon robots and automation, but the increased sales in areas beyond automotive is very encouraging for companies in automation.
Secondly, while these figures do not specifically cover machine vision technologies, robots and imaging technologies often go hand in hand. Industrial robots rely on machine vision to “see.” Vision-guided robot applications include pick-and-place, machine tending, assembly, and more. While “blind” robots still exist in industrial settings, adding machine vision into a setup increases flexibility and — in general — allows the systems to accomplish even more.
This is all to say that, while there have been some recent setbacks in the upward trajectory of the machine vision and imaging market, the technologies will only become more important over time. And eventually, the statistics should get back onto an upward trend. This will benefit not only those of us in the automation space but companies and individuals of all types as we navigate our way past the turbulent waters of the last year-plus.