Kyle Voosen, marketing director, National Instruments
How did you come to be part of the imaging/machine vision industry?
One of my early projects while attending Uni in the late 90s was to create a Where’s Waldo finder (Where’s Wally in the UK). I was using a text-based tool to perform some matrix manipulations and my algorithm took about 20 minutes to accurately find Waldo each time. During my senior design project, I was introduced to National Instruments LabView and I decided to rewrite my little Where’s Waldo program. It took me about 90 minutes to recreate the whole thing and the resulting LabView VI could find Waldo in less than three seconds.
After graduation I began working for National Instruments (NI), initially answering technical support calls. Looking back on it, building an imaging system was much more difficult in those days. Computers were relatively slow, digital camera connectivity standards didn’t exist and most engineers didn’t know what they were getting themselves in to. We’ve come a long way since then. I’ve spend the last 11 years at NI working with our imaging and machine vision teams. Currently, I also represent NI on the board of the Automated Imaging Association.
How do you convince customers that they need machine vision?
NI is lucky to be more than just a vision company; 35,000 companies come to us every year with a measurement or control problem to be solved and we consider machine vision to be one of several techniques that our customers can use in their applications. This also means our imaging tools are used in a wide variety of applications, from isolating circulating tumour cells to guiding autonomous ground vehicles through an urban environment. When we recommend a vision system to a customer, they have confidence that we have the breadth, experience and knowledge to point them toward the best solution.
What role does Europe have in the development of machine vision?
More than anywhere else in the world, European organisations have embraced machine vision as a valuable tool to automate production lines and increase quality. Because machine vision is already so well established, the customers we work with in Europe tend to have very complex challenges. Fortunately there is a strong commitment to research and development in Europe, which includes some of the world’s leading university programmes as well as hundreds of innovative companies. We try to give our customers access to this intellectual property through LabView libraries and interfaces. For instance, the 3D Shape Analysis Library from Aqsense is now available via LabView.
What do you see as the major growth sectors?
It’s no secret that intelligent traffic systems and warehouse logistics have been nice growth areas for traditional machine vision vendors. Looking forward, the days of selling high-margin barcode readers are probably numbered and the fastest growth for machine vision will come from outside the factory floor. One example that comes to mind is mobile robotics. Robots are coming out from behind their safety cages and are on the verge of entering many aspects of our day-to-day lives, from service robots wandering the halls of our local hospitals to autonomous tractors ploughing the fields. Nearly every autonomous mobile robot requires sophisticated imaging capabilities, from obstacle avoidance to visual simultaneous localisation and mapping. In the next decade, the number of vision systems used by autonomous robots should eclipse the number of systems used by fixed-base, robot arms.
What do you see as the most important technological challenges facing the industry?
We know how good an imaging system can be, because we humans are amazing processors of image data. Most machine vision systems, however, still require a skilled developer to program a string of algorithms, set detailed regions of interest and account for minute changes in lighting. If we could provide imaging software that could perceive a space, understand what it was seeing and tolerate changes in environmental conditions, then there would be no end to the number of applications for machine vision. This is not a camera, lens or lighting limitation; rather it’s a limitation in software and processing power.
What do you see as being the most significant commercial change in the industry during the years ahead?
Hands-down the biggest commercial challenge to the machine vision industry is the explosion of freely available, consumer-grade imaging tools. The 8 Megapixel camera that came with your smart phone isn’t that bad, and it’s only going to get better. Perhaps a little closer to home, how does an engineer justify paying £3,000 for a Data Matrix code reader when his boss’s smart phone does it for £0.99? For several decades, the machine vision industry has been pushing the imaging technology envelope. Soon it may be the vision companies that are playing catch-up to consumer technologies.