IEN: How do you define the difference between machine vision sensors and discrete sensors?
Blackwell: Machine vision sensors deal with images rather than discrete inputs such as force, strain, load, or temperature. As a result, vision sensors typically provide an order of magnitude more information than photo-eyes or ultrasonic transducers, and are suitable for a broader range of applications such as accurate gauging, guidance, identification, and inspection of objects.
While special-purpose sensors applied in industrial equipment require some configuration, the focus for users primarily remains in selection of the appropriate device for a particular job. In contrast, machine vision sensors are general-purpose devices suitable for a wide variety of operations. Consequently, setup and development require the application of vision tools to the acquired image in order to reliably accomplish a specific task.
IEN: What issues confront machine vision? How can they be resolved?
Blackwell: Ease of use is the main issue. In the past, users were concerned with the setup of a single vision sensor. Now users need to configure and manage an entire network of vision sensors. Cognex has already done its due diligence on the setup and configuration side by combining a spreadsheet interface with menu-based development environments.
The next step, to be unveiled as In-Sight Explorer at the RIA''s 2003 Robots & Vision show on June 3, will extend this user environment to include management and monitoring utilities. For example, the new In-Sight Explorer will not only make initial setup and application development easier, but it will reduce the total cost of ownership by minimizing the amount of resources required to maintain and monitor vision-sensor networks.
Miniaturization has also been a big issue. Making things smaller with fewer components improves reliability and reduces costs. Consequently, vision sensor suppliers continue to decrease the size and weight of cameras for use on moving equipment, or for embedding more deeply into machines. Machine vision suppliers are also doing a better job of integration. By packing an entire vision system with Ethernet connectivity into the space of a camera, the Cognex In-Sight 1000 and 4000 Series products are more modular and cost effective. As a result, some customers have hundreds of these vision sensors throughout their factories.
IEN: Has the use of vision systems spread beyond traditional end-use markets, and if so, why?
Blackwell: Yes. Though Cognex machine vision technology cut its teeth in the semiconductor and electronics industries, as our vision technology matured and became more reliable and robust, we developed new markets such as automotive, packaging, and document processing, to name a few. Moreover, by leveraging processing power increases, we can bring more of the high-performance vision tools from our PC-based systems over to the In-Sight sensor platform, which is easier and less expensive to implement in many cases.
The key drivers include greater line speeds, production flexibility, and product traceability. Human inspection and intervention can''t keep up across all industries where line speeds are on the rise. The need for production flexibility for more frequent changeovers has increased in many industries because the trend is to make a number of different products on a single production line. Greater product variability increases the need to error-proof the production process.
For example, in the automotive industry where multiple car models are built on a single line, our tire and wheel identification tools ensure that the correct tires and wheels are on the right vehicles. In other markets traceability is driving the need for 2D codes, which is making the general use of machine vision in code reading more prevalent.
IEN: What innovations are in store for users in the areas of: color, intelligent sensing, high-end edge detection, MEMS/sensing clusters, nanotechnology?
Blackwell: There will be more intelligent devices at the low end. We will continue to see a migration of high-end tools such as edge detection into the lower end. Machine vision users should watch for innovations in the area of easy-to-use color classification tools.
IEN: Will wireless play a role in sensing? In machine vision?
Blackwell: Wireless probably won''t play a significant role in machine vision other than for operator interface applications such as portable display technologies. While the need for wireless technology in hand-held type code readers is obvious, whether or not wireless will be viable for point-to-point image data transmission on a moving robotic arm remains to be seen.
IEN: How can sensitive data be protected?
Blackwell: Indeed, security is a key issue that must be resolved before wireless technology is more widely adopted in industrial applications.
IEN: How are software and equipment being improved?
Blackwell: Users can anticipate more ease-of-use improvements on the software side, as well as improved vision tool robustness to cope with more complex scenes and object variations. Improvements in industrial robustness are also on the horizon. Vision sensors will be more readily available with industrial connectors, washdown enclosures, and more flexible cables that hold up better in industrial applications.
IEN: How are sensing and machine vision being used in e-manufacturing?
Blackwell: In-Sight machine vision sensors with Ethernet connectivity can be linked to every level of the enterprise. Consequently, uplinked information such as good part count, number of rejects, and other data become available to various stakeholders within the enterprise. Information can also be downlinked from the enterprise level to the vision system. For instance, CAD data imports are used in defining inspection application models, and for training vision systems on new products. Similarly, for OCV applications, serial numbers, lot codes, and other information can be brought in to define the character strings that need verification during production.
IEN: Can sensing be part of a lean/flexible manufacturing solution? How about machine vision?
Blackwell: Machine vision is a key enabling technology for lean/flexible manufacturing. A vision-guided robot, for instance, can reduce the use of fixtures and hard tooling in picking or palletizing applications. In addition, vision sensors can help automate product changeovers in assembly and inspection operations