Imaging technology is playing an increasingly important role, for traditional industry sectors like manufacturing as well as emerging segments outside of the factory. Companies are looking to automate their operations. Lower costs have enabled customers to use camera-based technologies for the very first time. Vision systems have become more capable than single-purpose devices.

Alex Shikany, director of market analysis at the Association for Advancing Automation (A3).
As part of this month’s Imaging Technology Camera Directory & Guide, we asked experts from the Association for Advancing Automation, technology supplier analyst firm VDC Research, and technology market intelligence analyst company Tractica to talk about emerging camera capabilities and what machine vision customers can expect in 2015. The questions were posed separately to each of the analysts.

Imaging Technology: What is driving machine vision today?

Alex Shikany, director of market analysis at the Association for Advancing Automation (A3), Ann Arbor, MI: [Robotics and automation] are going hand in hand more often now. Companies are looking to give eyes to their system because it’s more cost-feasible for them to do so, and the capabilities of doing that really open up a lot of possibilities for them.

Richa Gupta, senior AutoID analyst at VDC Research, Natick, MA: There’s a lot of continued interest from manufacturing units in installing and deploying machine vision to automate their everyday processes. Robotics is certainly generating a lot of interest, as well as taking imaging technology beyond the manufacturing shop floor to the logistics environment. There’s just a whole host of applications that we never knew possible.

Richa Gupta, senior AutoID analyst at VDC Research.
Alex Shikany: Some of the main factors driving machine vision are really expanding outside the factory. You look at the mature industry inside the factory: process automation, inspection, measurement. Those are still very important to this industry, but what’s driving growth are when the companies are exploring areas like unmanned aerial vehicles or robotics or life sciences. Those vertical markets are really driving the increase in the market in recent years.

Imaging Technology: How has cost feasibility impacted the use of imaging technology?

Richa Gupta: The sharp decline in sensor prices has really opened up the market for their use and its application across the board. These machine vision cameras and systems are now being deployed for anything from damaged goods documentation to intelligent traffic systems to traditional inspection types of applications. You name it; cameras are being used.

Alex Shikany: The cost feasibility is opening up doors for new customers. Companies that couldn’t afford [machine vision] or were hesitant to implement the system are now starting to open their eyes. There’s more capability in each system now, and companies can offer it for less. You see that happening with robotics. Robots, in general, are costing less, which is opening up a lot of doors. That’s putting pressure, on an application-specific basis, on machine vision companies to offer their functionality at a reasonable price.

Imaging Technology: Where have you seen the impact of machine vision beyond the manufacturing floor?

Anand Joshi, senior analyst at Tractica
Richa Gupta: The volume of goods that are moving through logistics warehouses and distribution centers today is massive. It is something that manual labor alone is not really equipped to handle. A lot of these parcel and courier companies – like FedEx, UPS, and DHL – are relying on camera technology to help them automate their solutions, drive operational efficiencies, generate additional enhancements, and generate additional revenues, while keeping their costs low.

Anand Joshi, senior analyst at Tractica, Boulder, CO: On the industrial side: robotic technology will become self-guiding. Amazon, for example, acquired Kiva Systems. Kiva Systems has robots that can go from one end of the factory to the other and locate the package, bring it down, and deliver it to the appropriate shipping container.

Richa Gupta: I think it’s very interesting to really keep track of the amount of dollars that companies like FedEx, UPS, Alibaba, and Amazon are investing in robotics and image-based technology. Within the warehouses and distribution centers, I think that is probably the industry that is going to grow the fastest when it comes to machine vision adoption.

Imaging Technology: Is machine vision already pretty well embedded into the logistics process?

Richa Gupta: It used to be all laser scanners that were being used for over-the-belt scanning types of applications in these logistics environments. Now the laser scanners are being replaced by imaging solutions, especially with the tier-1 organizations that have a lot of money to spend. The camera-based imaging solutions are being tightly integrated into the material handling and overall warehouse automation solution. [The technologies are being used] not just for bar code scanning and bar code reading, but also in support of a variety of applications beyond track and trace.

Imaging Technology: What do customers expect from their cameras today?

A Kiva Systems robot moves product in Amazon's eighth-generation fulfillment center. (Image Credit: Business Wire)
Richa Gupta: Customers are so used to their iPhones, their Samsung tablets, and their iPads; they’re looking for devices that can support more than one application at a time and are not necessarily purpose-built. Camera technology now can be used for bar code identification, image capture, signature capture, damaged good documentation, and a variety of other applications that a purpose-built device cannot necessarily be put to use for. So they’re getting more bang for their buck essentially, which is really helping drive investments in this particular technology.

Anand Joshi: What used to happen with part inspection and QA [tasks]: There were all these fixed systems. So you would have one system that would only examine, say, nuts and bolts. Or PCB inspection, for example. You would need to feed a pattern into the system and then you could just inspect for that pattern. Now this new system is able to learn. You can train the systems to recognize new patterns and new objects. One system can be programmed to do multiple things.

Imaging Technology: What are new, emerging camera capabilities? What do you see trending in 2015 with camera technology?

Richa Gupta: A lot of the R&D efforts for organizations today are centered on enhancing the inherent power of the cameras, insuring that the megapixels are higher so that you’re no longer using a 2- or 3- megapixel camera on your machine vision system. There are companies that are launching industrial cameras that are up to 12 megapixels now, which is a big deal for an industrial setting. You’re talking about higher resolution, greater image quality, and you’re talking about leveraging these industrial cameras not just for black or white applications but also for color applications as well.

Anand Joshi: The resolution is improving. In the last 5 years, we’ve gone from 1 megapixel to 15-20 megapixels today. The low light response is improving. You’re able to get a better picture under less light. The whole processing side has gotten better. You have a lot more CPU, and a lot more RAM at a cheaper price.

Richa Gupta: There’s talk of pairing cameras, using cameras along with laser-sensing technology so that you can not only capture the image of a package, for example, but also capture the dimensions; this is a key requirement in logistics types of environments. That’s certainly something that you’re hearing a lot about from the camera space and from vendors that build these solutions.

Alex Shikany: Aside from your typical increased resolution, faster speeds, faster frames per second, and those types of hardware advances, I think the ability for a system to be embedded in different areas is really an interesting development. You’re seeing sensors being used for eye-tracking. Arthroscopic cameras are being used in surgeries and for exploring the human body.

Imaging Technology: What kinds of 3D imaging applications are possible?

Anand Joshi: The self-driving car, for example, from Google already has [3D imaging cameras]. They have multiple sensors on multiple locations, and they are actively gathering the data and essentially making decisions given the scenery around them or moving objects, including pedestrians. It’s mature enough to the extent that we feel ready to grant a driver’s license to Google’s car. We’re going to see a lot of applications come into the picture: automotive, industrial, consumer side, augmented reality, and such -- essentially computer vision technology extended to superimpose multiple images. In medical, for example, you’ll see 3D extraction of the heart. You can have four cameras looking at different directions, and then you can rotate the whole image.

A 3D image of the heart is shown projected onto a large screen. The simulation model can be used to test and improve medical devices. (Image Credit: Dassault Systèmes)
Alex Shikany: I don’t think the industry has quite settled on exactly what its main purpose is going to be. Certainly there are applications that beg for 3D reconstruction: inspection or augmented reality applications, for example. Even thermal is now starting to investigate the use of 3D imaging. The question is: What is it going to be in industry? For manufacturing, what is it going to be? That hasn’t been decided yet. There’s still a lot of hesitation because of the cost.

Anand Joshi: There will be more 3D extraction of images. You will see multiple cameras and the use of supplemental sensors like infrared, radar, and laser. What’s going to happen in the future is that we’re going to see these cameras used just as human eyes are. You’ll have two cameras extracting a 3D image. Or in automobile and other areas, you’ll have multiple sensors which will gather additional information, and the results are going to be a lot more accurate and lead to a fascinating set of applications.

Alex Shikany: Typically 3D vision systems are more expensive because of the capabilities. They’re more processingintensive so it takes more power on the back end to process a 3D system, as well as reconstruct the image. Your software needs to be more sophisticated. People are hesitant. A lot of times I’ll hear: “A 2D system is just fine for what I need to do right now.” 3D imaging is a niche right now, but it is growing and people are definitely intrigued.

For more information, visit the Association for Advancing Automation ( ), VDC Research ( ), and Tractica ( ).