How Routing Cleanup Buys Back Hours You Already Paid For

Most shops are carrying capacity they already own inside their own routings, in the form of wait gaps nobody revisited, alternate machines nobody referenced, and settings entered once for a reason that no longer holds. EDGEBIC by User Solutions ranks those findings in the BOR Optimization Suggestions report, ordered by the hours a change would recover. It suggests only. What it removes is the hardest part of routing cleanup, which is knowing which of four hundred routings to open first.

This post covers the recovered-hours outcome. It sits under the EDGEBIC results guide. For running the report, see how to run the optimization suggestions report.

Why Routings Rot Quietly

A routing is written once, usually under time pressure, often by someone estimating rather than measuring. Then it works. Nobody revisits a routing that works, because a routing that works produces no complaint.

Meanwhile the shop changes. A second machine arrives. A fixture is improved. A cure time is shortened by a better coating. A queue value entered as a safety margin during a bad quarter stays in place for six years. None of those changes reach the routing, because nothing forces them to and nobody is assigned to look.

So the plan quietly reserves time and capacity the plant no longer needs, and the cost shows up as lead time and as an overloaded station rather than as an obvious error. That is what makes it hard to find by hand: it is not wrong, it is stale.

The Seven Rules

The report ranks findings across seven rules, and each maps to a different kind of staleness.

Ranked by recoverable hours, the list usually has a long head and a short tail: two or three findings account for most of the available gain, and the rest are worth an afternoon at most.

The Two That Usually Pay First

Alternates configured but never used. This is the one that reliably surprises people. A shop buys a second machine, sets it up as an alternate once, and never references it on the routing steps that matter. The engine has no route to it, so it absorbs nothing. The primary reads CRITICAL, the second reads IDLE, and everyone concludes the shop needs more capacity. The fix is a routing edit. See the capacity you already have but cannot see.

Excessive wait gaps. Queue time is the most common place a safety margin gets frozen into a routing. Consider one step:

That 240 hours is not machine capacity, it is lead time: elapsed days on the quote that the process never used. Multiply it across a dozen routings and it is often the largest single contributor to a lead time nobody could explain. See manufacturing lead time reduction for the wider set of levers.

It Suggests, You Decide

The report deliberately stops at suggesting, and the reason is worth understanding rather than working around.

About half of what it flags is real staleness. The other half is a physical constraint the report cannot see: an eight-hour gap before paint that is a genuine cure, a one-per-day flag on a station where a single fixture really does prevent two jobs a day, a serial sequence that cannot be parallel because the operator has to be present for both steps.

The test is simple. Ask what the setting is protecting. If someone can name the physical reason and the number attached to it, leave it and write the reason into the routing so the next person does not spend an afternoon rediscovering it. If the answer is that it has always been there, you have a candidate.

Then change one routing, reschedule, and read the completion dates before touching the next. Changing eight routings and rescheduling once tells you the total effect and nothing about which change produced it.

Working the List Without Breaking Anything

A sensible cycle looks like this:

1. Run the report after a bulk import, a major routing edit, or on a quarterly rhythm.
2. Take the top finding only.
3. Test the reason. Name the constraint or accept it is stale.
4. Edit the routing and record why in the routing itself.
5. Reschedule and compare completion dates and the constraint's utilization rating against what you noted before.
6. Repeat with the next finding.

Two guardrails belong in that loop. Run the Scheduler Anomalies report after routing edits, since it runs 22 automated checks across seven categories and catches the data problems an edit can introduce. And remember that completed work is never moved by a reschedule, so testing a routing change on live jobs does not disturb the record of what already happened. See how the anomaly report keeps a bad schedule off the floor.

What This Is Not

It is not an optimizer. The suggestions report reads your routings and ranks candidate changes. It does not search for a better schedule, and it makes no claim about optimality. The scheduling optimizer is a separate capability with its own guarantees. See the business case for risk-free optimization.

It does not measure your process. It reads what the routings say. If a routing's hours are wrong, the suggestions are computed against wrong hours, and the fix for that is variance data rather than this report. See how variance data fixes the routing that was always wrong.

It cannot tell a cure time from a habit. That is the human half, and it is the half that decides whether the exercise creates value or creates scrap.

Recovered hours are not automatically output. Removing six hours of queue on a step that was never the constraint shortens the quote and changes nothing about throughput. The findings worth the most are the ones on or feeding the constraint. See how protecting the constraint lifts plant output.

It will not keep routings clean on its own. Staleness returns, because shops keep changing. A quarterly run is the difference between finding six months of drift and finding six years of it, and six months of drift is a morning's work. For the underlying waste framing, see the seven wastes of lean manufacturing.

Want to see what your own routings are holding? Bring a routing export to a demo and we will rank the findings over your data.