1. Introduction
Since the beginning of the 21st century, FANUC has successfully developed intelligent robots equipped with vision and force sensors. Currently, they offer a wide range of products capable of handling weights from 2 kg to 1.2 tons. Subsequently, FANUC developed the "Robot Cell," a long-duration continuous machining system utilizing these intelligent robots. While automated machining using CNC machine tools is commonplace in the machining process, many ancillary tasks, such as loading and unloading materials on machining center fixtures, deburring, and cleaning, are still performed manually. The Robot Cell, using intelligent robots, not only automates these tasks but also became the first in the world to achieve 720 hours of continuous machining. Currently, 13 such Robot Cells operate in FANUC's factories. The Robot Cell uses two control devices: a CNC and a robot controller. We are currently developing a function that allows both control devices to display their status independently, further promoting the integration of machine tools and robots.
I would like to take this opportunity to briefly introduce the main components of a robot unit. The "intelligence" in industrial intelligent robots does not refer to having the same thinking ability as humans, but rather to enabling the machine to possess operational skills comparable to those of a skilled worker. Robots without intelligence, that is, ordinary repetitive action robots, can indeed improve the level of automation in factories and reduce production costs. And once the robot's actions are taught, it can correctly repeat the actions, playing a significant role in stabilizing product quality. However, the problem is that humans must teach the robot every minute detail of the actions, which is very time-consuming. Furthermore, ordinary robots require specialized peripheral equipment such as workpiece feeding devices, or many workpiece positioning fixtures, which increases equipment investment costs. In addition, humans still need to pre-position the workpieces correctly—such simple daily tasks often involve humans assisting the robot's automated work. To solve this problem, we developed an intelligent robot equipped with visual sensors that function like eyes and force sensors that function like hands, and based on this, we developed a robot unit.
2. Robot Unit
(1) The Development Process of Long-Term Continuous Machining Systems For machining systems, minimizing initial investment and maximizing continuous operation time are crucial to reducing processing costs. In the 1980s, our first-generation machining system employed a method of loading multiple workpieces onto a traveling fixture with multiple loading surfaces, and then placing the fixture on an automatic feeder on an exchange table to supply material to the machining center, thus achieving 24-hour continuous machining. The second-generation machining system in the 1990s combined 4 to 6 machining centers with an automated warehouse capable of accommodating even more traveling fixtures, achieving 60 hours of continuous machining. This meant unmanned machining could be performed from Friday evening to Monday morning. Entering the 21st century, we developed the third-generation machining system, the "Robot Unit," using intelligent robots, achieving 720 hours of continuous machining per month.
(2) Composition of the robot unit The robot unit consists of a machining center, a robot for loading and unloading workpieces, a robot for deburring and cleaning workpieces, a pallet truck for placing the accompanying fixtures, and a material warehouse.
(3) The vision sensor of the robot for robotized workpiece loading and unloading operations can detect the position and rotation angle of the workpieces randomly placed on the tray. In this way, the robot can use a robotic arm to grasp the workpieces. In order to grasp a variety of workpieces, the robotic arm uses servo motors to drive the opening and closing of the fingers. In this way, the robot program can freely change the opening and closing degree of the fingers according to the size of the workpiece, and freely change the grasping force according to the weight of the workpiece.
To correctly position the gripped workpiece in the fixture, it's necessary to correct for gripping deviations caused by dimensional errors in the casting itself. The workpiece gripped by the robot is then measured in position and angle by another intelligent robot using a vision sensor capable of measuring the object's three-dimensional position and angle. Based on the measurement results, the deviation between the workpiece's gripping position during motion teaching and the actual gripping position can be calculated. When the robot positions the workpiece on the fixture, accurate positioning is achieved by correcting for the gripping deviation. Finally, as the robot loads the workpiece onto the fixture, its axes are controlled in a special way to make the robot's movements smooth in a specified Cartesian coordinate system. This allows the robot to automatically adjust the workpiece's position and angle based on the fixture's surface condition. This method, called Soft Float (follow-up function), enables high-precision workpiece loading onto the fixture.
It has the advantages of making full use of the space and being easy to operate, making it a system that is easy to introduce.
(4) Large Robot Large machining centers are generally used to process heavy workpieces. Therefore, FANUC has developed a "large robot" with a maximum handling capacity of 1.2 tons to meet the processing needs of heavy workpieces. The large robot can easily grab the casting bed of a small machine tool and load it into the fixture of the machining center.
(5) Two-way status display on the CNC and robot controller teaching panels: To enable operators who only know how to operate the robot or only know how to operate the CNC to operate the robot unit, FANUC has developed functions that can display the CNC status on the robot teaching panel screen and the robot status on the CNC screen. In the future, we will continue to develop functions that integrate the two, further promote the integration of CNC and robot, and take improving operability as our future goal.
3. Ending
We believe that robotic cells are highly competitive machining systems capable of significantly reducing processing costs through long-term continuous machining. I would be extremely pleased if this information provides useful data for readers.
