plied by a third party and was not only inferior, but incompatible with the rest of the system.)
The AGVS includes 16 identical carts, 53 basically identical pallets, about 100 fixtures of either vertical or horizontal design, and about 150 subplates of various designs. Each cart carries a pallet with the work on it to a machine and transfers the pallet onto the machine's shuttle. (The shuttle rotates 180 degrees and can hold two pallets at a time—one that is being worked on and the other being loaded or unloaded.) Only one part, or job, is loaded on a pallet at a time. The load/unload workers build up the pallets by attaching a horizontal or vertical fixture to the pallet, depending on the job to be machined, and then bolting the most appropriate subplate onto the fixture. The part to be machined is then firmly bolted to the subplate. This is shown in Figure 4.
Through a tow-release mechanism, the AGVs follow a track of rails that run about 400 feet and include 30 zones for machine centers, load/unload stations, and accumulator (buffer) spaces. Only one cart is allowed in a zone at a time; if more than one are in a zone the computer shuts the system down. As a designated cart reaches its assigned zone, a limit switch embedded in the floor is thrown by the cart to tell the computer that the cart is in position. On occasion the carts get hung up in their position or in their load transfers to the shuttles and need a nudge from the machine operator to get them going again.
The entire system is controlled through two computers that transmit information back and forth between themselves. One computer serves as master and DNC (direct numerical control) computer for the machine tools and the other operates as a slave for the FMS system. The master handles the machine tool part program data, passes it to the machining centers, and interfaces with the machine operators.
The slave handles data files (part, route, tool, pallet), status files (machines, pallets, carts, etc.), and the AGVS. It checks and updates the files as activities are completed and the carts are moved from zone to zone. It interfaces with the AGVS through a digital multiplexor bay activated by relays, solenoids, and limit switches. It also interfaces with the load/unload crew at their stations.
Adjacent to the FMS is a dedicated tool crib. Medina maintains a total of over 10,000 different tooling numbers throughout the plant, though not all of them are used on the FMS. Here, tools are preset and staged for the next production run. The tools are delivered to the machine and manually loaded onto each machine's tool chain, which can store 59 tools at a time.
The FMS is currently programmed to machine 65 different parts in batch sizes ranging from 1 to 300. Cycle times on these parts range from 40 minutes for simple operations to 160 minutes for the most complex. Inspection is accomplished by programmed delays in the system for gauging purposes, which commonly take about five minutes each.
The FMS is currently operated by a four-person crew (two operators and two load/unload people) in the day shift and three in the evening and graveyard shifts (only one operator). Their time, and other over
Figure 4: AGV cart and system at machine center.
Figure 4: AGV cart and system at machine center.
head costs such as maintenance and setup, are allocated through standards for machining time. Although current plant burden rates are 250 percent variable and 300 percent fixed, the FMS is charged at 290 percent variable and 406 percent fixed due its minimal direct labor content.
One of the increasingly important advantages of the FMS to Medina has been its ability to replace worn out lines that made replacement parts for obsolete products. The FMS has also been particularly good for low volume startups on new products, especially prototypes.
However, the Medina FMS was one of the very early FMS installations and some aspects of the system tended to become bottlenecks or problems as the system throughput increased. About four years previously, Medina began to experience trouble with some of the electronic components. Fortunately, they had excellent electrical maintenance people who could solve the problems. But with age, some accuracy problems also began to crop up. These became even more serious with the increased stress on quality by Medina's customers.
A list of the increasing maintenance problems as detailed by plant maintenance is given in Table 1. In essence, severe availability and compatibility problems are being experienced with replacement parts, primarily for two reasons. First, with the passage of time there are fewer suppliers and compatible replacement parts available. As noted in the table, some of the manufacturers of the FMS parts are not even in business anymore. And, of course, it is less worthwhile to keep old replacement parts for any piece of equipment, particularly one-of-a-kind equipment.
Secondly, as the FMS parts age, more and more components start wearing out and requiring replacement. Also, new or slightly different replacement parts place a heavier load on the existing parts until the new parts have "worn in," thus leading to even more failures of existing marginal parts.
Plant maintenance's comments about the system are typified by the following: "The printer is operating only by the grace of God. At present, it ejects three
•fable 1. FMS Evaluation by Plant Maintenance_
Category I: FMS Serviceability_
1. FMS Computers: Serious Incompatibility problems with new computer boards because of this being a prototype system. The computers are no longer serviced by the manufacturer and expensive third-party service Is thus required.
2. 5MB Disk Drives: Obsolete; no longer serviced by manufacturer.
3. Digital Multiplexor Bay: Obsolete; no part availability; no longer serviced by manufacturer.
4. Hard Disk (for NC programs): Obsolete; no longer serviced by manufacturer.
5. Line Printer: Obsolete; manufacturer gone bankrupt.
6. Software: Obsolete; a prototype system no longer supported; written in an obsolete version of Fortran proprietary to OEM.
1- Relay Cabinets (10): Obsolete; parts totally unavailable; if one part fails an entire cabinet-full must be rebuilt.
Machine Centers: 14-years-old; wear in slides, ways, and ballscrews; cannot hold tolerances adequately to continue making two critical components on the FMS. 9- Track Relays (200): Becoming unreliable; replacements totally unavailable. Category |i; FMS Malfunctions_
1- FMS down three times in the last six weeks. CRT down six weeks last year.
3- Electronic malfunctions five times a week (on average), each taking about four hours to repair.
4- Mechanical malfunctions (table hang-ups, center of rotation problems, and shuttle fail-_ures. most commonly) require 25 hours a week in total to repair.
pages of blank paper between each printed sheet. We have not been able to repair this and replacing the printer is impractical due to FMS software compatibility." In conclusion, they note that upward compatibility of the FMS system was not maintained by the manufacturer and thus, individual elements cannot be continually replaced because they must electrically match the old components perfectly.
The retrofit proposal includes reworking all the machines and the software to bring them up-to-date. For example, plans are to rework the three-axis machines into state-of-the-art four-axis centers with a machine tool rapid travel speed increase from the current 150 inches-per-minute to 400. Further, the dual computer system would be upgraded to a single computer, such as a DEC VAX®, with system software and machine control drives designed by one of Medina's sister firms.
Dave and his team figure that the project will probably take almost two years to complete and cost about $3.5 million. (The original installation took over two years to get all the bugs out.) The project is divided into a hardware and a software/control system stage. The hardware stage, Phase (, runs 20 months and consists primarily of remanufacturing the eight machining centers and refurbishing the material handling system. The machines would be sent out two at a time for remanufacturing with a turnaround of approximately 20 weeks. This would allow the FMS to continue operating at Medina with a reduced operating capacity from the six remaining machines.
Phase II, the control system stage, consists of installing hardware and software and will run about 55 weeks, most of it concurrent with Phase I. The project would start in mid-1986 and finish in mid-1988 but over 90 percent of the expense would occur in 1987.
The proposal identified the approximate cost savings the retrofit would bring. Based on the current $12 million worth of parts put through the FMS, annual savings would accrue in four areas:
Indirect repair labor 30,000
Outside support 15,000
3. Scrap and rework (35% reduction): 10,000
4. Reduced part subcontracting: 100,000
If the proposal is not funded, then additional outsourcing will be required at an annual expense of $200,000, plus a one-time, up-front tooling expense of $75,000.
However, there are some non-cost elements of the retrofit project that Forrest Daley also pointed out in the proposal. Perhaps chief among these was Medina's "plant mission" as Geartrain's supplier of low to mid-volume strategic parts such as product variations, prototypes, service parts, and general "cats and dogs." Strategic parts were defined in the proposal as "key proprietary components integral to our differential carrier and drive axle assemblies requiring close, in-house control of quality." Outsourcing these strategic parts would result in Geartrain losing control over both the proprietary manufacturing process and the required quality of the key parts.
As the proposal concluded: "This program is critical in supporting Medina's continuing role within the Corporation as a low-cost, highly flexible supplier of quality components. We must take some action if Medina is to continue over the long term manufacturing components on this line that will meet our world-class quality criterion."
The problem facing Dave was the amount of work that needed to be done in so many areas of technology. Although the plant was not particularly busy at the present time, no one seemed to realize that this is exactly when the staff are the busiest. The workloads are always countercyclic due to a number of reasons. First, when production volumes are down, there are more runs of smaller batches which thus require more tooling, more setup, and more of everything that is of a staff nature.
Second, when production, and thus profit, is lower, there is more need for cost improvement and other such "staff" types of programs (value analysis, zero defects, and so on). For example, with the reduced workload in the plant they now were instituting an SQC (statistical quality control) program. The FMS project was being undertaken at this time because throughput demand was lower than previously. For example, it was now running at only 92 percent utilization of two-shift capacity (one shift = 1800 labor hours/year).
Yet, because profits are down, Dave can't ask for additional people to help him accomplish the work.
Was this article helpful?
What you need to know about… Project Management Made Easy! Project management consists of more than just a large building project and can encompass small projects as well. No matter what the size of your project, you need to have some sort of project management. How you manage your project has everything to do with its outcome.