Shop-Floor Search Time: Log Retrieval Failures and Retest

Measure shop-floor searching as a retrieval task with a defined start and finish. Record what someone needed, why it was hard to retrieve and whether the correct item or information was found. Retest the same task after one change before estimating any wider benefit.

A person can spend six minutes looking for a gauge even when the gauge is in the right place. The location label may be unfamiliar, the requested identity may be incomplete or access may depend on someone who is unavailable. Moving the gauge will solve only some of those problems.

Choose one retrieval task

A supervisor and a person doing the work should agree the scope. Start when repeated searching interrupts a recurring job. Select a specific request, such as finding the released setup sheet for one operation or locating one identified tool. Bring the request, expected answer, current storage or document location and access rules.

Observe with the team's knowledge and follow site access rules. Keep the record about the task, not a ranking of individual workers. Start timing when the person begins retrieval; stop when the correct item or information is available for use, or when the attempt is abandoned. Record unsuccessful attempts too.

Separate search time from unrelated work

Illustrative scenario: five observed attempts contain 3, 5, 2, 6 and 4 minutes of searching. The total is 20 person-minutes, and the mean is four minutes per attempt. In the six-minute case, an additional two-minute unrelated conversation occurred; the elapsed clock interval was eight minutes, while the recorded search component remains six.

If two people search at once for three minutes, elapsed time is three minutes and labour effort is six person-minutes. Keep those measures separate. Neither proves that a machine stopped or that wages could be reduced. Record a production block independently if it occurred.

The five attempts describe this small observed set. Do not multiply them across every operator and shift as though they were representative. Include the task, conditions, count and unsuccessful attempts when sharing the result.

Fix the failure you observed

The Lean Enterprise Institute's 5S explanation includes arranging items for use and sustaining workplace order. For a missing-location problem, an agreed home and a readable identity may be enough. An access problem needs the authorised access route. An item already in use needs a reservation decision rather than another label.

For document retrieval, make the released revision distinguishable from superseded copies. A fast search that returns the wrong instruction fails. For physical items, check that the new home preserves handling and storage requirements; convenience alone does not justify moving controlled equipment.

Retest correctness before speed

Run comparable retrievals with the same definition and relevant operating conditions. Retain the original observations and record changed conditions, such as a different shift or someone helping. Acceptance requires the correct result, permitted access and a clear return or update process. Compare the distribution of times and unsuccessful attempts as well as the average.

If a label helps only the person who designed it, ask another authorised user to retrieve the item. If the location is empty, the record should expose the current holder or exception owner. Review the location after a short operating period to see whether returns follow the agreed rule.

We support related operational records and role-based form access in Optiwise. Configure a retrieval issue log that links the task to the relevant record and responsible role. For delays in transfer between departments, use a separate handoff observation; its start and finish describe a different problem.

Search-study questions

Should abandoned searches count?

Yes. Record their duration and unresolved need separately from successful retrievals. Excluding them can make an ineffective process appear fast.

Can reduced searching be reported as cash savings?

Only with separate evidence of the cost change. A shorter observed retrieval establishes time change under those conditions, not a payroll or output result.

Bring one repeated retrieval problem to an Optiwise workflow demo.