Robotics Case Studies
- This case study is filed under:
Kuntz Electroplating Automated Wheel Polishing System
by Cathy Powell
FANUC America Corporation
Kuntz Electroplating Inc. of Kitchener, Ontario is one of the world’s largest independent OEM chrome plating suppliers, producing millions of plated automotive wheels, bumpers and other components annually. Kuntz is a family owned business which has been in operation for over 50 years, employs over 1,000 people at peak production, and is an approved supplier to the Big Three, Harley Davidson and many others.
In recent years, a trend in new wheel styles often includes deeply contoured spokes. Traditional fixed automation polishing and buffing machines cannot reach the deeply recessed areas between the spokes of the wheel. To accomplish the polishing operation, hundreds of finishing workers are required to work in a 24/6 manual operation.
The manual finishers polish the wheels with sandpaper cartridge rolls mounted on die grinders. The subsequent chrome plating operation requires that the finish on the wheels be of exceptionally high quality. Any areas that are not blended or smoothed properly will be very noticeable once the plating operation is complete. Finding, training, and retaining hundreds of manual finishing personnel became a daunting task. Manual polishing is tedious and monotonous, the average new worker lasted less than a month. Also, these polishing operations accounted for more than half of the entire company’s repetitive strain injuries.
Because of these difficulties, it became apparent to Kuntz Management that some type of automation was required. After conducting a survey of available technology, it was determined that Kuntz would have to develop their own automated process for wheel finishing. Kuntz has extensive experience in designing and implementing fixed automation for automotive wheel finishing, but it was determined that fixed automation would still not be effective for the newer wheel styles with deeply contoured spokes. Therefore, a completely new approach was needed for this application.
Kuntz contacted FANUC Robotics at an early stage of the project in search of an of-the-shelf solution. Dr. Laxmi Musunur, manager of the Material Removal group at FANUC Robotics, had completed many robotic finishing and deburring systems, but had never attempted to polish aluminum wheels with such stringent quality requirements.
After some discussion, FANUC Robotics was able to help Kuntz select the appropriate robot and process equipment and several things became evident. A variable speed (up to 6,000 rpm), high torque motor, capable of automatic media change and a compliant tool would be needed. PushCorp, Inc., in Dallas, Texas, was chosen to provide a critical component in the end-of-arm tooling. Their active force compliant tools would provide the compliance and accurate forces necessary to duplicate the human touch of the manual finishing operation.
Description of Process or Solution
Initial test results were very promising, and Kuntz decided to install twenty M-710i robots. This was a very ambitious project with ten cells of two robots each. The cell design featured two inverted robots mounted on an elaborate steel superstructure. The superstructure was situated above a large turntable capable of fixturing ten wheels.
Following start-up and de-bug, the first of these cells began producing wheels in mid-1999, with the last cell being completed about one year later. At this point, a very difficult issue arose. Robot paths for the first three wheel styles were programmed by manual lead-by-teach methods, and each wheel path program was taking eight to 10 weeks to create.
The lengthy programming time was due to the extreme complexity of the wheels, one of which had 15 spokes. There was also a considerable learning curve to overcome trying to understand the fine points of the human operator’s techniques and reproduce them adequately. The goal was for the robots to achieve the same smooth, blended finish characteristic of wheels that had been manually polished.
To address this problem, it was decided that some type of off-line robot path programming software was needed. To acquire this capability, Kuntz purchased a Krypton RODYM-6D Robot Measurement System and contracted the services of Al Knasinski of Logic Systems Development Inc. This combination of equipment and personnel allowed the development of an Off-Line Programming (OLP) solution based on using a 3-D CAD model of the wheel. These techniques promised to give better control over all of the parameters, such as force head pressure, spindle speed, travel speed, and blending paths to produce smoother, more consistent paths in less time. The system could also be used to calibrate the robots and easily transfer programs from one cell to another.
Work began in earnest utilizing the robot measurement and programming system to develop robot programs for the new model-year wheels. Unfortunately, this effort brought to light a significant shortcoming in the cell design that was not apparent with the original 1998 model-year wheels.
The 1998 wheel styles were shaped such that the sandpaper cartridge orientation relative to the wheels did not deviate much more than about 10° from vertical. However, in 2000, Kuntz began working on developing paths for the 2001 model-year wheels. It soon became apparent that there was a fundamental problem with the cell layout. The 2001 wheels had a significantly different spoke shape that required that the sandpaper cartridge be able to reach positions at a much shallower angle relative to the wheel face.
Even after a costly conversion of the wheel fixtures to reposition the wheels, a number of other issues became apparent. The new robot path programs required much faster motions, which caused flexing in the robot support superstructure. This led to surface defects in the polishing when one robot’s motion would cause the other to vibrate. Also, the larger motions required by the new wheel styles often forced the robots into singularities.
After making a concerted, but ultimately futile effort to make the original cell design work for the new wheels, Kuntz decided to reconfigure the cells into a totally new design, which was referred to internally as the “Second Generation.” In the new cell layout, the robots were mounted conventionally, upright on a pedestal. In front of each robot are two programmable wheel mounting platters, which are the robot’s seventh and eighth auxiliary axes. FANUC Robotics came to Kuntz’s assistance by supplying one of their M-6i robots.
As the robot polishes, the wheel is rotated about its own axis, allowing coordinated motion with the robot. Because the wheel is continually repositioned, the robot works within the dexterous part of the work envelope. While the robot polishes a wheel on one auxiliary rotary axis, the operator can unload and load the other unit, permitting continuous operation. The two auxiliary axis units are separated by safety fencing; each has pressure safety mats and a complete set of operator controls. There is an elaborate system of safety interlocks on the auxiliary rotary axis units and on the robot to ensure that the operator can safely load one unit while the robot works on the other. This is not a trivial consideration, since the unit the operator is loading is still “live” and under control of the robot controller throughout the duration of the operator’s exposure!
Throughout the effort to get the original cells performing properly and development on the Second-Generation concept, there were other significant developments underway. The first was a business decision by Kuntz Electroplating Inc. to form a new division called Kuntz Logic Systems Inc. (KLS) The new division was intended to develop off-line programming solutions, robot metrology and integrated systems. The business unit was formed by purchasing the assets of Logic Systems Development Inc. from their majority owner, Krypton Electronic Engineering of Belgium. The asset purchase was completed in late 2000, and KLS officially began operations on January 1, 2001.
The basic concept of Off-Line Programming (OLP) is not a new one, but as many people in the industry are aware, the Achilles’ heel of OLP is that it has been very difficult, if not impossible, to achieve a simulation-created robot path that requires no real-world touch-up. Because the polishing paths necessary for automotive wheels are among the most difficult robot programs, the task was extremely challenging.
To polish automotive wheels, there are numerous complex and interrelated factors to coordinate. Of primary importance is that the working axis of the PushCorp compliant tool must always be normal to the surface at the point of contact. Also, the motor travel speeds, compliant tool forces and motor RPM are in a constantly varying balancing act dependent upon the local surface geometry. So, if any one factor must change, the others must be automatically adjusted to suit. Acceleration and deceleration must be carefully controlled at all times. Initial contact and lift-off points must be feathered to avoid polishing discontinuities. Also, wear patterns on the sandpaper cartridges must be controlled for a long, even polishing life. The sandpaper cartridge lead and toe-in angles must be constantly adjusted for smooth blending of polishing marks. In short, it is nearly impossible to develop such complex paths by manual lead-to-teach methods. With some styles of polishing media, there is a considerable change in tool diameter as the media wears down, which also has to be compensated for by the program.
Another complicating factor is the inconsistency of the lug nut hole locations relative to the wheel spokes. The wheels are positioned by these holes; but, according to the manufacturer’s specifications, the placement of these holes can vary rotationally by +/- 1°. On an 18 inch (457 mm) diameter rim, this leads to an error of almost 1/4 inch (8 mm) near the outer edge of the wheel. This can cause major problems with robotic polishing. KLS subsequently developed a proprietary method for automatically detecting and compensating for this error that does not require the expense of a vision system.
Lastly, the OLP system has to be capable of quickly creating different versions of any one wheel polishing path. This is necessary to be able to cope with the varying surface quality of the incoming raw cast aluminum wheels. When the surface conditions deteriorate, the robot operator needs to have the option of adding additional polishing paths, or even different grit sizes of sandpaper to the base routine from the robot’s control panel. This means that the programs must be capable of adapting on-the-fly to changing requirements without having to create or load new programs.
The process of starting a completely new business division and subsequently developing an OLP package is a very complex and time-consuming task for a chrome plating supplier. However, at the time, Kuntz management felt that this was the best solution. Several companies supply OLP packages, but each suffer from various limitations that severely limited their usefulness to Kuntz Electroplating.
The task of developing OLP solutions for polishing wheels starts by loading a 3-D CAD model of the wheel into Robcad™. Next, the wheels are subdivided into various portions based on their geometry, and detailed subroutines are developed to handle the particular requirements of each kind of geometry. Then, for each wheel, a master program assembles the subroutines in the correct order depending on the number of passes selected by the operator for each different grit size. At this point, a certain level of intelligence is added to the system to semi-automate the process of assembling the master program. This application is referred to as APG, or Automatic Path Generation.
Another very significant piece of software developed for this project is a new interface between the simulation software and FANUC RCS (Robot Controller Software), which KLS calls the “LS_RCS Interface.” All of the new KLS software runs as a ROSE (Robcad Open System Environment) function in Robcad™, which is a robot simulation software package produced by Tecnomatix Technologies, Inc. Like all robot simulation software packages, Robcad contains a robot motion planner; however, a robot path generated with simulation software may not run the same way when it is downloaded to a robot. This is because the real robot controller software is different in many respects. Our solution was to develop an interface that allows the programmer to select the actual FANUC RCS (made available separately by FANUC Robotics) to do all the motion planning. In fact, this new interface allows the full use of every function contained within the RCS, which immediately opens up a much wider range of programming tools to a simulation engineer.
There are two other essential elements used to ensure a touch-up free download to the robot. The first is another KLS application - an OLP Interface that ensures 100 percent round-load integrity. This means that there is absolutely no loss or degradation of data in a robot path transferred from simulation software to the robot controller or vice versa. The second essential element is that the target robot polishing cells are fully calibrated using a RODYM-6D™ Robot Measurement System produced by Krypton Electronic Engineering of Belgium. The calibration procedure compensates for any dimensional irregularities between the simulated robot cell and the real one. One additional piece of KLS software called VitualTeachPendant, which is very helpful but not necessarily essential, is a teach pendant emulator, which lets the OLP operator view, modify and run the robot path in simulation exactly as on a real teach pendant.
KLS has now reached the point where a polishing path for a new wheel can be developed by a skilled programmer in approximately one to two weeks using all of the tools developed to date, including testing and optimization. This compares very favorably to the eight to 10 weeks that it took for lead-to-teach paths in the original polishing cells. In addition, the automatically generated paths are far superior in terms of smoothness of motion and overall improvement of wheel quality.
Obviously, this approach to OLP has broad implications aside from polishing wheels. For example, both FANUC Robotics and a major automaker are using the new RCS Interface as part of a software package supplied by KLS for off-line programming of paint robots.
of a software package supplied by KLS for off-line programming of paint robots.
There is additional work that is presently at the research stage, including a major R&D project at the University of Waterloo, which should further reduce that interval. The U of W project was set up with the assistance of a number of companies. PushCorp of Dallas, Texas provided their latest Active Force Head and Spindle. FANUC Robotics Canada, Ltd. supplied an M-710i robot and controller. Kuntz Electroplating Inc. provided all of the remaining hardware, and is donating a complete Krypton RODYM-6D Robot Measurement System. Just as the new RCS Interface has broad implications beyond aluminum wheels, it is expected that the work being done in the university R&D project will move OLP to the next level. One of the first areas of research is to develop the means to automate the task of optimizing the robot paths. This would eliminate a great deal of the trial and error in testing various robot polishing paths.
By late 2001, Kuntz built a prototype Second Generation Pilot Cell consisting of two robots and four auxiliary axis units. The new system produces extremely consistent work and has exhibited none of the problems that plagued the first generation design. The quality of the output is nearly indistinguishable from that of the best human polisher. After an extended trial, the remaining robots will be reinstalled in the new configuration.
The ability to quickly program complex automotive wheels provides Kuntz Electroplating a unique manufacturing advantage. These cells can obviously be used to run continuously on large production runs, but they also offer a very advantageous short-run capability. Many automakers are now looking at more frequent changes of trim components, such as wheels, as a means of offering special editions and for keeping their overall product offerings fresh. Also, by diversifying its customer base beyond the Big Three, Kuntz expects to considerably increase the number of short runs.
With large product runs of 10,000 wheels or more per week, the manual operators are able to hone their proficiency, resulting in greater throughput and improved efficiency. However, with short runs, some as low as 500 or less per week, this level of proficiency never develops. If short-run wheels can be polished on a robot, then the quality level and throughput will not depend on achieving and maintaining operator proficiency.
To this end, Kuntz has been using their Robot Lab development cell to test and develop the means to incorporate the other types of polishing media in addition to the sandpaper cartridge. These media may include small flap wheels, brushes and sandpaper spinners of various styles. The goal is to be able to eliminate all hand polishing for any wheel style which requires these processes. The first wheel to be programmed for robotic polishing using all of these types of media as required was a wheel for a major automaker's SUV, which is a very large, complex surface area to be polished. The polishing quality achieved by the robotic process was found to equal or exceed that of the best human polishers, but at only 1/2 to 1/3 of their typical time.
Labor Savings: Yes
Injury/Risk Reduction: Yes
ANSI/RIA R15.06-1992: Yes
Total Robots Installed: 20
Date Installed: 1999
Shifts Per Day: 3