Read an Excerpt
HyGain-Telex: Analysis for JIT Production
Distinction between preventive maintenance and setup
Frequency of delivery
Statistical process control
Total preventive maintenance
Simplifying the schedule
Partnership with customer
The HyGain-Telex plant in Lincoln, Nebraska, manufactures antennas. It currently has a U.S. Army contract for Model X32 antennas. The contract requires a production rate of two hundred Model X32s per day. The contract quantity may be changed quarterly.
Chris Piper, the foreman, is collecting data for a JIT project. Piper has selected the X32 antenna base (not the whip part of the antenna, which is fairly simple) for the JIT project. Exhibit 1-1 is a photograph of the base.
Manufacture of the X32
There are several stages of manufacture for the X32 base, which is a cylinder 6 inches in diameter and 10 inches high. These are the basic production processes, and their standard times, with which Piper was concerned:
* Mold the Lexan plastic base. Some holes are molded into the base by use of core plugs. 2.50 minutes.
* Drill and tap (eight operations). A dozen more holes are drilled; half of the drilled and molded-in holes are tapped, and half are installed with "helicoils" self-threaded inserts (a rather old technology). Seven drill or tap operations taking from 0.12 to 1.02 minutes; installing helicoils, 1.82 minutes.
* Assemble (epoxy) as "birdcage" (ferrite core, coaxial cable, etc.) inside the Lexan base. 1.78 minutes. (Note: The birdcage is produced as a subassembly, going through twelve operations.)
* Foam the assembly. 2.61 minutes.
* Paint. 1.82 minutes.
Piper felt that the place to start was between drill-and-tap and assembly. Drill-and-tap ran one shift, and assembly usually ran two shifts. Piper asked L. G. Smith, the industrial engineer, to find out the flow distance between processes, especially those two processes. Smith scaled off the distances on the factory blueprints and came up with a total flow distance of 1,296 feet, which breaks down as follows: from mold to drill-and-tap, 192 feet; from drill-and-tap to tank assembly, 144 feet; from assembly to paint, 480 feet; and from paint to final prep, 480 feet.
Piper wanted to be sure. "Are those prints current?" he asked. Smith assured him that they were. Just to make sure, Piper got a tape measure and checked some of the distances; they were indeed correct.
For flow-time data, Piper went to Raul Nieves, the scheduler. Nieves pointed out that the flow time from molding to final prep had been "as short as about five days for a few lots, but we are quoting six weeks to marketing." Piper asked Nieves to come up with some sort of average. Nieves did so by putting pieces of colored tape on a few molded bases from several lots over the space of three weeks. The average flow time, found by noting how long it took for the taped units to get to final prep, was seven weeks. One week of that was the flow time from the start of drill-and-tap to assembly.
Question 1. What is the ratio of actual production lead time (or flow-time) to work content time from the start of drill-and-tap to final prep?
Nieves also provided Piper with scheduling and unit-load data. Scheduling released work packets in lot quantities of 2,000. Drilled and tapped bases were forwarded to assembly by forklift truck, in wire-bound pallets holding about 400 bases. In other words, about five forklift trips were required to move one "packet-release" quantity to assembly.
At this point Piper called a meeting. Smith and Nieves were there, along with Karen Jones, manager of quality assurance; Bob Crane, an inspector; Doug Atkins, a drill press operator; and Ellie Olson, an assembler. Piper announced that the purpose of the meeting was to "brainstorm what can and maybe can't be done to reduce WIP and flow time" between drill-and-tap and assembly. Piper explained that the purpose was to improve and not look for blame. In that spirit, "please speak frankly."
Piper's first question was directed to Atkins: "Doug, there's no setup time on the drill press that you use for the X32 it's a dedicated tool, right?" Atkins said that it was.
"How about up time on the drill press? Is it reliable?" asked Piper. Atkins replied that the drill press itself was fine but that the tapping head with spindles in the taps were a problem sometimes: "They break or the bushings loosen," which results in off-center taps or a marred surface around the outside. "Then I have to call maintenance to make adjustments or replace the head."
"About how many hours per month are you down waiting for them to make those adjustments or replacements, Doug?" Atkins estimated about five hours.
Ellie Olson was next. "Ellie, do you have any problems with the bases? Quality problems or running out of bases?" Ellie said that sometimes she did have to wait for the fork truck to bring another wire-bound; she estimated six hours of wait time per month.
The quality problems were the biggest headache, Olson felt, and she looked at Bob Crane, the inspector, for corroboration. Crane agreed that the defect rates were high, especially cracks and fractures around the helicoil inserts. Some, "maybe 5 percent," they thought, were minor defects that Crane or Olson let pass. Crane had figures on how many were defective but repairable and defective-scrapped: 2 percent repaired, 4 percent scrapped.
Karen Jones, quality manager, pointed out that their customer, the Army, had been rejecting an average of 7 percent in recent months. "I believe that the majority of the problems can be traced back to drill-and-tap," she stated.
Piper then asked if anyone knew what level of work-in-process (WIP) there was of bases. Nieves said he had just made a rough count; there were six wire-bounds full at drill-and-tap and eight and a half full at assembly.
Question 2. If fifteen direct labor employees are involved in the production of the Lexan base, what is the ratio of pieces in process to people who could work on them?
At this point the group began brainstorming on J IT opportunities. Some of the options they discussed:
1. Setup reduction (adjust and replace spindles/bushings) on the drill presses. To this suggestion, everyone nodded their heads, but no one commented pro or con.
2. Cut transit quantities. Nieves (scheduler) protested: "The fork truck drivers would be making more trips."
3. Adopt kanban. Nieves liked the idea.
ard4. Use process control charts in drill-and-tap. Everyone thought it was about time to do some of this.
5. Adopt total preventive maintenance. This was Piper's (the foreman's) idea. The others showed little reaction; they seemed not to know what that meant.
6. Put in conveyors. Smith (the I.E.) offered that one; nobody challenged the idea.
7. Slash the buffer stock. Nieves suggested this, pointing out that inventory counting was a headache anyway. Olson was indignant: "I run out of bases too often the way it is."
8. Get rid of the packet-release quantities. Smith suggested this but admitted that he did not know what kind of scheduling might replace the packet-release system.
9. Bring the design engineers in to come up with a better design of the base. Everyone smiled and nodded vigorously.
10. Expand the size of the task force (which they were calling themselves by that time), including a customer (Army) representative. This was Jones's suggestion, which was met by a couple of favorable nods.
11. Move a drill press into the assembly department. This was Smith's idea. Crane (inspector) said that "if we do that I won't have to inspect the bases and I'm not complaining; it's a boring job."
The meeting broke up with plenty of ideas but no decisions.
Question 3. What should be done? Should all the ideas be implemented? None of them? A different set? What order? To what extent? What time period? What guidance and direction? Discuss each of the eleven options that came out in the brainstorming session.
Copyright © 1987 by Schonberger & Associates, Inc.