Industrial Robotic Vision Systems

Robotic vision systems can assist in increasing plant productivity, safety and quality. Vision systems offer a critical use of technology for robotic applications. With increased use of vision technology in manufacturing, production economies of scale have helped reduce the prices of some of the critical components used in integration of vision systems, making them more available and affordable.

Dramatic improvements in robotic performance can be achieved through the use of these integrated vision systems, locating required parts, welding accurately, and inspecting final assembly and replacement of parts. In addition, vision systems include capabilities for parts sortation, flexible parts feeding, presence/absence, error proofing, visual servoing and environmental perception. When used in dispensing applications, they also ensure that liquids are being dispensed to the right locations.

2D Vision Systems

For standard 2D vision systems, cameras can be mounted up, down, horizontal, or at angles within the robotic workspace. Alternatively, cameras can also be permanently affixed above the robotic workspace or mounted flexibly moving with end-of- arm tooling. Flexible mounting provides higher resolution imagery in addition to allowing for six degrees of freedom within the workspace, but there are advantages to both mounting configurations. These cameras are available in IP67 rating and provide additional protection against dust and liquids in industrial environments, designed to withstand shock and vibration during robotic uptime.

3D Vision Systems

For applications that require more sophisticated vision systems, 3D cameras area available to provide three-dimensional vision across the X, Y and Z axes. 3D vision assists with correct placement of randomly place parts. Certain 3D software can even be “taught,” and are capable of handling extreme part variability with ease, therefore making random depalletization, bin picking and deformable objects ideal applications. While 3D vision software offers more features than that of the 2D variety, most application needs can be met with 2D vision systems.

With spatial-type vision software, 3D vision can be achieved through the use of a standard webcam; however, calibration may be required when using multiple images. Standard CAD files of parts may be used to track data after being imported into the software. Additionally, spatial vision software is accurate up to 1/10th of a pixel, making it incredibly reliable.

Robotic Vision Techniques

Images captured by 3D cameras are compared to those stored within the software template. The two popular techniques used for robotic vision applications include PBVS (Position-Based Visual Servoing) and IBVS (Image-Based Visual Servoing). When the PBVS method is in use, object depth is calculated, so adequate lighting is especially important for these applications. Vision tools of varying performance and price are available to assist in quick and easy programing of vision applications, depending on your vision system’s requirements.

Augmented Sensing Technologies

Vision systems can also make use of additional augmented sensing technologies, allowing for the accurate detection of size variation and parts location. When the vision system camera captures an image and sends it to the processor, the processor has the ability to convert that image to a digital signal which is then interpreted by the software.

Once this occurs, the vision software can then sort through geometry, color, barcode, markings and other pre-defined characteristics.

Additionally, cameras are controlled by high-speed Ethernet connections, which provide enough bandwidth for image processing. GigE Vision, for example, is a popular standard used for industrial cameras, which runs on UDP protocol and PoE (Power-Over-Ethernet), simplifying the amount of cabling used. With the emergence of mobile technology, mobile vision applications have been developed, and can be connected through wireless Ethernet as well.

Benefits of Robotic Vision Systems

Robotic vision systems have the ability to track and trace parts from start to final assembly of a product. This means that any defective parts can be traced back to final assembly.

Moreover, better analytics can be generated by tying the vision system information to your company’s MES and/or ERP systems for SKU, shift and vendor analysis. These analytics can provide real-time data to your server or cloud computing system, tying together production, order processing and inventory management.

Notably, the barcodes scanned with these systems can be compared to databases, eliminating the selling and production of counterfeit goods.

These characteristics lead to increased quality of a final product, as a result of rejection of defective parts, increased throughput of parts manufactured, increased productivity through reduced usage of resources and increased output. Part processing cycle times are marginally impacted through the use of vision systems, but increased productivity results in a decrease in the payback period for an installed system. Accurate distance measurement with automatic focus, increased mean time between failures (MTBF) and reduced setup time with robot vision systems are also benefits to vision system usage.

One of the most important benefits of robotic vision systems is that they can be used in hazardous locations, reducing industrial accidents in an industrial environment. It also reduces costs associated with overuse of materials, making production processes leaner.

To learn more about implementing robotic vision systems, contact a Motion Ai Specialist today.