Dan McCarthy, contributing editor for the Association for Advancing Automation in the US, reports on training programmes available for the modern machine vision engineer
The automation industry is booming. But if qualified job candidates are out there, they don’t appear to hear the noise. Skilled help is hard to find, and according to a recent study by Deloitte and The Manufacturing Institute, the hiring crunch won’t ease up any time soon. The study forecasts that 4.6 million manufacturing jobs will be created in the US over the next decade, but more than 2.4 million of those jobs will remain unfilled by 2028. Currently, manufacturing executives report an average of 94 days to recruit engineers and researchers.
Several factors are contributing to this shortage, including the looming retirement of baby boomers combined with a perception of manufacturing as a dirty, thankless job by the generations who might replace them. The latter factor, at least, appears to be changing. The 2019 L2L Manufacturing Index, which measures the American public’s perceptions of US manufacturing, found that, compared to the general population, young Americans between 18 and 22 years of age are 12 per cent less likely to view the manufacturing industry as in decline and 7 per cent more likely to consider making manufacturing a career path. That begs the question: how does one study for this path?
The way forward
The machine vision industry, which plays a vital role in automated manufacturing, shares in the sector’s struggle for talent. It is also contributing to the solution by seeking and finding help from a growing list of partners in academia — as demonstrated by AIA’s Vision Research and Academia page (www.visiononline.org/vision/vision-research). Today, students have more options than ever before when it comes to pursuing an engineering school or advanced degree in computational imaging.
But while degree programmes provide general knowledge, many cannot offer hands-on experience. Savvy machine vision companies try to address this by partnering with nearby colleges, universities, and vocational schools. Edmund Optics, for example, works closely with the Rochester Institute of Technology.
Even so, it is one thing to learn how a barcode scanning system works. It is another thing entirely to adapt and build one for a particular application, and yet another to be able to troubleshoot the system when it goes sideways. Each of these tasks represents the progressively higher levels of knowledge and expertise required for barcode readers – a comparatively common machine vision application. Yet, like any highly technical discipline, machine vision is constantly evolving new technologies and applications, often rapidly, and even veteran engineers can find themselves falling behind without regularly training and continuing their education.
Adapting machine learning to vision systems, for example, is a specialisation that has risen to the top of priorities for employers such as Andy Long, CEO of automation integrator Cyth Systems, who says the acceleration in the demand for machine learning is staggering. The US Bureau of Labor Statistics agrees. It reports that demand for machine learning professionals is expected to increase 11 per cent by 2024.
Another example is embedded vision, which lies at the intersection between machine vision and embedded electronics. Embedded vision is emerging as a discipline all of its own, with increasing adoption in consumer electronics, automotive and medical applications.
Vision-guided robotics, which requires knowledge in vision, software and advanced motion controls, is another growing field that demands more specialised expertise beyond conventional machine vision system assembly.
Keeping pace with technology
Machine vision curricula in academia and industry are adapting to address these industry trends. Courses in machine learning and vision-guided robotics, for example, are being added to a growing number of machine vision courses in colleges and continuing edge programmes.
The machine vision industry is following suit, as evidenced by AIA’s Certified Vision Professional (CVP) programme. For years, CVP has provided a baseline education for machine vision engineers, according to AIA vice president Alex Shikany, who said: ‘A lot of big names in the machine vision industry use CVP certification as a level set. They put new employees through the programme, so they come out with a baseline, broad understanding of the technology.’
Offering two tiers of certification, CVP attracts a broad spectrum of attendees, including recent college graduates, as well as seasoned professionals looking for a career move. Consequently, it is subject to frequent adaptation to keep pace with industry trends.
The programme is undergoing one such review now, according to Robert Huschka, AIA’s director of education strategies, who says that the aim is to ensure the curriculum meets industry needs. ‘We looked at the courses and saw that attention to machine vision was peppered throughout,’ Huschka said. ‘But we decided we really needed a course focused entirely on machine learning.’
AIA is also reviewing its content for embedded vision and autonomous navigation, to see whether or not these topics need more independent coverage.
Technology is not all that is driving such curriculum reviews. The nature and tools of learning are as well. Another key goal in CVP’s review processes is to examine how to best deliver the content.
‘We’re focused on the experience,’ Shikany said. ‘So, we’re taking the opportunity during this process to interview people and companies who have used the CVP programme, so we can come out on the other side with a 21st century learning experience. We’re sensitive to the fact that people learn in different ways, and there are new tools available now.’
For example, AIA is exploring the idea of adding online training, expanding the pool of instructors, and looking for ways to make the courses more experiential. The goal is to make the coursework more attractive to more students, while delivering an education that meets the needs of industry today.
Attracting more candidates
Adapting training programmes to a constantly evolving industry is a perpetual process. But before you can train talent, you must first attract it. The challenge for the machine vision industry is finding ways to make machine vision engaging. According to James Gardiner, business development manager for Metaphase Technologies, the robotics industry has an edge over machine vision in this regard. ‘With robotics, the students can physically touch the solution and see it working. For machine vision, unfortunately, we’re more of a black box,’ he said.
Fortunately, vision is an increasingly pervasive technology, in industries as diverse as automotive, consumer electronics, medicine and manufacturing. Partnering with academic and industry certification programmes serving these sectors will continue to be an important strategy for vision companies.
There is no quick or enduring answer to the machine vision industry’s need for qualified talent, unless constant adaptation is the answer. As academia’s growing attention to imaging and its specialities demonstrates, there is awareness of the demand, and we can expect both demand and awareness to grow as machine vision solutions become a more common part of consumer experiences. But smart companies won’t wait to capitalise on training as an important tool for recruitment, retention and competitiveness.
Write for us
Have you attended a machine vision training programme and would like to write about the experience? Are there vision topics where training is missing? Please write for us! Email: greg.blackman@europascience.com.
