Читать книгу Surviving the Spare Parts Crisis - Joel Levitt - Страница 12

CHAPTER 3

MODERN PRODUCTION GIVES US THE TOOLS FOR MAINTENANCE PLANNING

Interchangeable parts are an old concept traceable back to standardized ship building in Carthage during the first Punic war. In the United States, Eli Whitney was credited with the idea when he demonstrated to Congress the advantages of interchangeability. He brought ten rifles to Congress built with identical parts; he disassembled the rifles and put the pieces into piles. He and his team reassembled the weapons from random parts in a few minutes.

Building any machine in the era of interchangeable parts requires a Bill of Materials. In addition, you need the tools as well as the competency needed for the assembly. These three elements — parts, tools, and competency —became the core of MRP (Material Requirements Planning). For this purpose, tools are defined in the widest way possible.

THE MIRACLE OF MODERN PRODUCTION

Think about the complexity of building a modern automobile. Thousands of parts have to be sequenced into the assembly areas with rigorous timing. Each operation requires specific tools that have specific set-ups and people with the correct skills (correct competency).

On the assembly line, the parts, sub-assemblies, and assemblies are received just in time for the specific car they are destined for. Very limited stock is stored on or near the line. To make matters more complicated still, the most modern assembly lines can run different models without setup or changeover.

Behind this complex dance is a huge infrastructure micro-managing the supply chain. The scope is large with thousands of SKUs, precise timing, and multi-modality inbound shipping including ship, rail, truck, and airplane. All, by the way, have different lead times, vendors, and development cycles.

Some say this array is the ultimate expression of the modern manufacturing world. Keep in mind that behind the assembly line is an army of people and systems to make sure part X appears at time Y in position Z (all the parts, materials, and consumables needed to assemble the part into the car).

TIMING AND TRUST ARE EVERYTHING

To make this mechanism operate, OEMs (Original Equipment Manufacturers) issue forecasts to the vendors with annual estimates of quantities and estimates of delivery dates. The vendors place orders consistent with these numbers for their own raw materials. The whole system is based on the vendors being able to deliver what is needed when it is needed. It is essential that the OEMs and the vendors trust each other so that they can share detailed sales information to sharpen the forecast.

Because the forecast is ultimately an estimate, albeit an educated one, the actual orders that come in are different from what was anticipated. Vendors must be able to change quantities, models, and colors on a moment’s notice. They have to take up the slack in the system and deliver just what is requested when it is requested. The system has to deal with schedule changes, but within a limited and (mostly) controlled manner.

WHERE DID THIS ALL COME FROM? LET’S REVISIT FORD AND HIS MODEL T

As discussed earlier, 100 years ago Henry Ford pioneered some of the concepts of having critical inventory in motion. His staff had iron ore cars from Minnesota and coal cars from West Virginia constantly in motion to arrive at the River Rouge plant complex in Dearborn, Michigan, just in time to feed the blast furnaces needed to produce the hot-rolled steel mill used to shape the car bodies.

(A note to readers with an interest in industrial history: over a decade ago, tours of the River Rouge facility resumed in cooperation with The Henry Ford Museum, as well as viewings of the assembly floor.)

The steel was formed and welded by Fisher Body in Detroit and coated with paint from DuPont in Delaware (any color you wanted as long as it was black). His revolutionary assembly lines were fed rubber tires from Akron, Ohio, as well as engines and transmissions all made up not quite just in time. All of these efforts were tracked and managed manually using systems and procedures that were state of the art in 1920. The Model T and later the Model A rolled off the line by the thousands (Figure 3-1).

Figure 3-1: The auto assembly line

WHAT IT TAKES DAY IN AND DAY OUT

To make even the simplest automobile, there has to be a precise convergence of skilled workers, machine, tools, each and every material on the BOM (Bill of Materials), consumables, and data sheets. If anything is missing, the line stops or the cars go off to rework queues. Either case rapidly becomes a disaster because of the speed of the assembly line.

Big maintenance lesson: To fix any machine requires the same precision coordination of these three elements: competent workers, all the tools, and all the materials.

When a production machine needs service for either preventive maintenance or breakdown, it needs, in precise timing,

• Skilled or competent workers

• Proper tools, machines

• Every material on the BOM, including consumables

If anything is missing, then either the job is not done or the maintenance workers improvise to make the repair as best as they can.

SAME LOGIC BUT DIFFERENT APPLICATION

In automobile production, there are a few thousand SKUs needed in large quantities to build the current crop of automobiles. The SKUs are forecast in advance, but the actual requirement can vary with production.

To repair and service machines, you have a universe of hundreds of thousands to millions of SKUs needed, with no way to forecast 60% of the requirements. (As many as 40% of the SKUs can be forecast, such as belts, bearings, etc.)

As mentioned earlier, preventive maintenance requirements are driven by an internal schedule and can be easily forecast. Delivery can (and should) be scheduled with the vendor. In fact, in stable production plants, the PM parts can be blanket ordered for a year with deliveries monthly.

A second category of parts that can sometimes be scheduled (which we previously called corrective maintenance parts) are repair parts where the machine is on some kind of effective surveillance such as PM-visual inspection, vibration analysis, or oil analysis. Using some kind of surveillance to predict when a component should be replaced is especially effective for long-lived assets where deterioration is relatively slow and consistent.

Not every kind of failure (e.g., a stone breaking a car windshield) is detectable far enough in advance that the BOM for the repair can be forecast. Beyond that, unfortunately, even with the best surveillance, machines will fail due to operator errors, problems with materials or tooling, and random causes.

REACTIVE VERSUS PROACTIVE PARTS REQUIREMENTS

In up-to-date manufacturing facilities, maintenance is 80% proactive and thereby planned, then scheduled, as measured by direct labor hours. The rest of the workload (20%) is reactive due to either machine breakdowns or do-it-now (DIN) type work that emerges and is performed that day. This 20% generates most of the problem spares because we never know when we will need them. Compounding the problem is the enormous number of spares needed if we want full coverage.

BUFFER STOCK

Modern supply chain management and JIT (Just-in-Time) supply strive to remove or reduce buffer stocks. Buffer stock is a very bad term in an up-to-date assembly line. But even those lines have buffers. Some say that, if there are no buffers on the line, the vendor will hold buffers because of the consequence of a mistake shutting down the assembly line. Even so, they might only hold eight hours of stock or more depending on the variability in the supply chain.

With machine repair, the buffer stock problem is significantly more complex. Some of the drivers of this complexity include:

• More SKUs

• More vendors

• Small purchases (less clout with the vendors)

• Wider range of ages and technologies of parts needed

• Inability to know exactly what parts will be needed until the asset is opened

• Obsolete parts needed for operating machines

• Less transparency of vendors

To deal with the complexity, the field has settled on keeping a stock of parts in a maintenance warehouse or storeroom. The alternative is massive redundancy so that the breakdown of any machine will not impact output.