Section D. Uptime Is Too Low


D. % Uptime Is Too Low

Overview

Processes are required to be generating entities (doing VA work) for 100% of the shift time. Any time they are not doing VA work can be considered downtime. Note that this is different from the definition typically used by maintenance personnel, which focuses on the availability of the process to do any work, rather than the process of actually doing work (and that work being all VA work).

This is very similar to the measure known as Utilization, which considers % work time, but does not consider the delineation of value-added (VA) versus non-value-added (NVA) work. For a detailed explanation of % Uptime, refer to "Overall Equipment Effectiveness (OEE)" in Chapter 7, "Tools."

Examples

  • Industrial. Uptime of a line, vessel, or plant

  • Healthcare. Surgery uptime (incision to close)

  • Service/Transactional. Uptime of a server or carrier

Measuring Performance

In considering uptime, typical primary measures would comprise

  • % VA Uptime (current shift pattern) measured as the percentage of the existing shift pattern on which the process is performing VA work. Note that VA work does not include

    • Load

    • Unload

    • Rework

    • Inspection

    • Planned process maintenance

    • Unplanned process maintenance

    • Breakdowns

    • Scheduled workforce breaks

    • Scheduled process downtime of any form

    • Changeover time

    • Setup time

Often it is useful to also consider the uptime relative to a 24-hour working week, multiple shifts, and so forth, and report these as separate numbers to show the true available capacity of the process, not only that which is limited by current staffing.

Tool Approach

If not already done in a previous step, then apply the sequence of tools as follows:

Focus should just be on measuring validitya sound operational definition and consistent measure of % Uptime versus a detailed investigation of Gage R&R. For more details see "MSAValidity" in Chapter 7. Determine an exhaustive set of categories of time (e.g., planned maintenance, unplanned maintenance, changeover, etc.). Having a category called "Other" tends to swallow up informative events and consequently any opportunity for improvement, so be as specific as possible.

Use the process operators to capture data for the time categories as defined in the preceding MSA. Ensure they have a clear understanding of the intent of the project to try to reduce the Hawthorne Effect[2] (the Champion might provide valuable input here).

One week's data typically provides a reasonable starting sample for all but the slowest cycling processes, but continue the data capture for long enough that the process time categories are all hit. Consider using historical data to represent large downtimes due to maintenance or forced shutdown.

Categorize the total (shift) time into uptime and all the possible downtime buckets using a Pareto Chart. Use the tool to focus the next steps of the project on the biggest sources of lost uptime. There might also be small sources that can be tackled very easily with a quick hit (i.e., negligible cost, dumb stuff). Take these opportunities immediately.

Options will be either

 

If the changeover for the whole process or for a single step is too long, go to Section O in this chapter.

If time is spent reworking, inspecting, scrapping, or reacting in any way to defective entities, it is a quality-related problem and should be addressed using the roadmap in Section C in this chapter to eliminate the source(s) of the defects.

If there is too much unplanned maintenance or the process simply breaks down on a regular basis, go to Section P in this chapter.

If there is too much planned maintenance or planned maintenance takes too long, go to Section N in this chapter.

This outcome will not be examined in detail in this book.

Consider using different staffing models, staggered breaks, and so on. Use the operators themselves to come up with the mechanisms to keep the process running during breaks. This helps ensure buy-in, which is typically the most difficult factor in these situations. Consider just keeping the bottleneck (the rate-determining step) in the process running during the breaksit will require a buffer of inventory ahead of the bottleneck to achieve this.


[2] Individual behaviors might be altered when they know they are being studied. First demonstrated in research at the Hawthorne Plant of the Western Electric Company, Cicero, Illinois (19271932).

For any other downtime category, unfortunately, you are on your own. Start with simple tools to map the problem or look to the Problem Categories in this chapter and Chapter 4, "Individual Step Process Problems," for the smaller problem to see if there are any guiding lights.




Lean Sigma(c) A Practitionaer's Guide
Lean Sigma: A Practitioners Guide
ISBN: 0132390787
EAN: 2147483647
Year: 2006
Pages: 138

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net