What’s happening here? An auto-industry supplier was faced with a sharp increase in orders for engine mount brackets when it was awarded a contract for an international market as well as the existing domestic market. Shipments fell behind schedule. Quality complaints required 200% visual checking by the manufacturer – once at the manufacturer and again at the customer’s site – and, still, defective parts got through. The added costs of quality, expedited freight, and overtime pushed production costs of these brackets into the red and to the top of the “hot list” to fix. Can production catch up? Can a complete solution be found to eliminate all defective parts
The Situation as observed on the shop floor:
The engine mount bracket is made from flat coil stock, which is rolled into cylinders 3 inches in diameter and 2 ½ inches long in the first press/operation. The bracket that now ships also to the international market has only one notch instead of two. In all other respects, the brackets are identical. The cylinders are then loaded manually into a second press, which forms a ½ inch flange on one end of the cylinder. The first operation runs five times faster than the second operation. One operator working one shift on machine No. 1 for the 245 normal work days in a year will produce almost exactly the 2 million parts required for both the domestic and international markets. The struggle is getting the second operation to run at the same speed. In order to do that, three shifts run the second machine, and another 1.5 shifts run on the weekend. As many as 55 metal containers are stacked along the wall with engine mount brackets from the first operation waiting to have the flange formed on them. Two operators of the second machine on each shift work to load the press with an unflanged part, then unload the flanged part onto a conveyor which drops into a shipping container. The containers of finished parts move to a quality hold area where the inspectors will check each part for slivers, flanges on the wrong end of the cylinder, no flange, open seams, wrong (domestic vs. international) part, and rust. A second inspection for the same defects is also being done at the customer’s facilities. Complicating the historically poor success rate for visual inspections is the minute difference between the right bracket and the wrong one.
The Findings
The manufacturer had made the too common, too tempting, always-seemingly logical mistake of trying to speed things up by adding more shifts or parallel production lines. The reasoning goes like this: If there’s a bottleneck on the freeway, it can be freed by opening more lanes. That makes sense on a freeway, but adding lines or shifts only adds costs and adds them exponentially. In this case, six operators worked round-the-clock shifts, and overtime shifts were added on the weekend! Worse from the customer’s standpoint, hopeless complexity and uncontrollable variables all caused defective parts. Some defects were attributable to the operators, some to the material, some to the equipment. Vision sensors were added for error proofing, and mirrors installed for better visibility. Dedicated locations and color-coded labels for the different parts were put in place. Other tooling issues, i.e., breaking bolts, sensors faulting out the press, two parts in the die, and other mechanical and electrical faults also occurred randomly. New untrained operators further confused attempts to isolate the cause of the machine faults and the defective parts. Problems occurred daily. The problems on off-shifts took longer to fix because maintenance and tooling people are dedicated almost solely to the first shift. Attempts were made to design automation for the second operation. The designs could neither achieve a reasonable payback nor a complete solution to making the 8,333 parts that had to be produced each 24 hours to meet the customer’s demand.
The Solution
A. Stout & Associates eliminated the entire second step by making it part of the first. All second-step labor was reassigned to profitable work elsewhere. The first-machine operator now produces everything by initiating a process that automatically completes itself. All the wasted time of producing first parts only to store them in barrels, retrieve them, load them into a second machine, unload them, reload them, etc., is now reduced to tenths of a second by use of a robot and pneumatic piston that load and unload the second operation at speeds approaching that of the first machine. The first machine rests momentarily every so often while the second machine finishes and is ready to receive more.
The Results
ROI on the robot was three months as measured in reassigned operator labor alone. Eliminating the work-in-process inventory and keeping the parts correctly oriented in the conveyer eliminated flanges formed on the wrong end of the part, missing flanges, and mixed-parts defect. The number of parts produced per hour increased from 216 to 1,028 pieces, meaning that in just one shift one operator produces 8,224 parts which will almost total the required 2 million annually. Revenue increased 5 times for every hour of operation.
Interested in what your asset could really be worth?
Contact me to learn more.