Recognition Robotics has developed a smaller, more cost effective robotic guidance system based on previous product requirements for a large automotive OEM. According to the Vision Systems Design report, Recognition robotics launched the Robeye Robot guidance program in 2012. The system uses an industrial camera connected to the robotic arm to capture part images to guide the correct location of the robot parts and to have an independent industrial computer to handle the image and communicate with the robot controller. Recently, a car customer needs to take up a smaller footprint, lower cost, and set up and operate simpler systems. Recognition Robotics then cooperated with Adlink to develop Robeye's improved version Robeye All-in-one (RAIO), which provides visual guidance for the robotic arm used to remove the body panel from the rack and place it on the vehicle. Raio equipped with Adlink NEON-1020 Smart camera. The camera uses AMS sensors Belgium (cmosis) 2 million-element CMV2000 CMOS image sensor, drawing element size up to 5.5 microns, frame rate up to fps. The Adlink industrial Smart camera carries the Intel (Intel) Atom 4 Core processor E3845 1.91GHz, 1 FPGA coprocessor to support Windows or Linux operating systems. Adlink says this architecture reduces the footprint of Robeye products and provides IP67-grade self-contained vision recognition and guidance systems. Adlink says Raio is a lightweight solution that can be directly mounted to the machine arm and, because it is a stand-alone system, can be remotely connected Raio and robotic controllers for physical teaching, setting and execution. With this capability, multiple Raio units can be programmed and monitored remotely on a manufacturing site network. Raio also uses recognition robotics visual algorithms, which handle images in a way similar to the human visual cortex. The system can recognize the parts from a single 2D image and transmit the X, Y, Z, Rx, ry and RZ position information of the part to the robot user architecture. With this data, the robot can offset the data at the current position of the part and update the pre-programmed path. Recognition Robotics, executive Simon Melikian, said the algorithms were developed based on the ability of human beings to perceive objects.
