Robots and motion systems are mainstream components in automotive, electronics, packaging, clinical, food and beverage, warehousing and general manufacturing applications. They perform operations such as pick and place, assembly, packaging and palletizing, screwdriving, labeling, gluing, dispensing, welding and polishing.
With the addition of a vision system, a robot is able to achieve a fuller level of autonomy, performing tasks with a greater level of speed and accuracy. A combination of a robot/vision system working together can perform higher-value operations than either would be able to accomplish alone. Rather than just manipulating an object, the robot, in conjunction with a vision system, is able to verify that it is working with the correct components in the first place, and can then verify that the operation was done completely and correctly through post-inspection.
When looking to integrate a robot and vision system together, a number of challenges need to be addressed to ensure success. Physical challenges such as camera size, weight, placement and mounting are especially important to pay attention to because the camera is often mounted on the end of a robot arm. Cabling is another area of concern. Making sure you have high-flex cables capable of millions of motion cycles are a must. You also want to ensure your camera capability. The camera must have the required horsepower and algorithm set, as well as flexible optics and lighting all capable of performing the required tasks. Before beginning any job, make sure you have addressed these areas of integration from the start.
In addition to the challenges outlined above, there are additional user interface and integration challenges to be aware of. First, the integration needs an ideal programming environment for achieving your desired results. An ideal programming environment allows programming and control of the robot and the vision system within one integrated environment. Second, calibration of the vision system to robot coordinates, whether in 2D or 3D space must be finalized. Finally, the robot and vision system must share common standard communication protocols so they can communicate with one another, as well as send and receive data with other systems on the factory floor—a key component of machine-to-machine communications, which underlies the Internet of Things (IoT).
Miniature, highly capable smart cameras are ideal for many applications that require the combination of a robot or motion system with a machine vision system. One example is Microscan Systems’ line of smart cameras called the MicroHawk MV. These cameras are extremely small and lightweight, and require just a single cable for power and I/O. They come equipped with built-in lighting, as well as built-in optics with fast programmable focus. Communications range from USB to serial to Ethernet, supporting major standards such as TCP/IP sockets, EtherNet/IP and Profinet.
A typical robot vision example application would be a smart camera mounted on the end of a robot arm tasked with verifying that all of the components on an assembly, such as an automobile engine, are present and correctly installed. The robot is used to fly the camera to different points around the engine, pointing the camera at each area of interest. Each inspection point is assigned a unique trigger. Upon receipt of that trigger, the camera takes a picture with the optimal sensor gain and exposure, as well as focus distance for that inspection. The camera then runs one or multiple tests ranging from reading of 1D or 2D codes on labels, to presence/absence, quality inspection, counting and gauging. With each unique trigger, a unique result is passed back to a database and stored against that engine’s serial number.
Another example involves a miniature smart camera working in conjunction with a multi-axis motion system, working collaboratively to assemble multiple parts. A cellphone is a good example of an assembly application where multiple small parts need to be assembled with very high accuracy. The robot steps one part of the assembly under the camera, presenting a view of multiple key features on the part. At each location, the camera finds the exact location of a specific feature on the part with micron resolution. The robot then steps a second part of the assembly under the camera, which again finds and records the locations of companion features on this second part. Because the camera is calibrated to the robot coordinates, the location of all points is now known in the robot coordinate space for both parts of the assembly. Calculations are then made using a rigid body fit algorithm with this point set to determine the exact x, y and angular offset of one part to the other. This offset is passed back to the robot, which then uses the data to assemble the parts together in exact detail.
There are numerous other examples where a miniature smart camera is the ideal choice based on form, fit and function—from surface-mount technology (SMT) assembly pick-and-place machines to desktop clinical analyzers, or when multiple cameras are required to be mounted very close together to cover the entire viewing area of a part. Additionally, because of the camera’s low cost and high performance, it can be mounted over a workspace to locate objects that the robot needs to find and pick up. A scenario for this type of application would be locating loaves of bread moving on a belt that need to be picked up and packaged, or finding parts in a bin.
Robots, when combined with a vision system, are highly capable of performing both sensing and manipulation tasks. Working together, the system can perform much higher-value operations than can be done with the robot alone. Miniature integrated smart cameras offer definite advantages over larger PC-based or traditional smart cameras because of their small size, light weight, variable sensing and programmable focus. They can be mounted out of the way on robot arms or be placed in very tight spaces. Microscan’s MicroHawk was designed specifically to tackle the challenges of integration with different robot applications.
To learn more about integrating miniature smart cameras into robotic applications, watch Microscan’s on-demand webinar at http://awgo.to/microscan.
For more information, visit Microscan at www.microscan.com.