Resource readiness is control, not a checklist

In regulated manufacturing, “resource ready” usually means three things at once: the equipment is in the right state, whatever enables it is still within validity, and the person who will run it is authorized. On paper, that becomes a start-of-shift checklist. In a spreadsheet, it becomes a green column. On the plant floor, the risk question is different: can the batch start when one of those conditions has already expired?

Good manufacturing practice treats equipment, utilities, and people as part of process control — not as room decoration. Cleaning status, calibration, maintenance, qualification, and training are not POP ornaments. They are preconditions for execution. The difference between a checklist and control is when the failure is found: before the start, or in the investigation after the deviation.

What a checklist actually buys

A well-filled checklist records intention. It asks whether someone looked. It does not guarantee that what was looked at is still true in the second the order starts.

Three limits show up early:

1. Validity is time, not a stamp. Calibration that was fine yesterday may be overdue today. Cleaning done on shift A may not cover the product on shift B. Training completed that quarter may have let the skill required for that step expire.
2. The checklist does not block. If the operator ticks and continues, the execution system never knows a “no” was hiding on a parallel sheet.
3. State changes after the signature. A maintenance intervention, a format change, or a utility alarm can invalidate readiness minutes after the batch’s opening “ready.”

That is why serious resource readiness looks less like a form and more like a latch: the resource only enters the flow when state, validity, and authorization line up — and it drops out when any of them falls.

A checklist proves someone asked. Control proves the process did not advance without the right answer.

What “ready” needs to mean

A useful resource-readiness model usually joins three axes on the same resource (line, scale, room, person, utensil):

• State — clean / dirty / in maintenance / under qualification / blocked. State without an owner and without an auditable transition becomes shift opinion.
• Validity — calibration, preventive maintenance, cleaning expiry, solution expiry, skill expiry. Without a date and an expiry rule, “valid” is only an adjective.
• Authorization — who may release the resource, who may operate it, who may override. An override without a trail is the shortcut inspection finds later.

When those axes live in separate systems — calibration in a CMMS, cleaning in the EBR, training in the LMS, status in the supervisor’s head — the plant reconciles at start time. Human reconciliation under pressure is exactly the failure mode the checklist pretends to have solved.

The silent failure mode

The worst case is not the resource that is clearly out. It is the “almost ok” resource operations decide to use “just this once.”

Concrete signals:

• Batch starts with calibration due the same day and no clear cut-off rule
• Cleaning signed for one product and mentally reused for the next
• An operator covering a post with an expired skill “because nobody else showed up”
• Maintenance open in parallel while the order stays “in production”
• Contamination or data deviations whose root cause is a resource used out of state — found in review, not at start

In those cases, readiness documentation existed. The barrier did not.

Impose readiness at the moment of control

The shift is the same one behind EBR, holding time, and dispensing: move control to the instant when damage can still be prevented.

In practice that asks for:

1. The start queries the resource, not memory. The order only advances if the resources tied to the step are in a compatible state and validity.
2. Expiry is an event, not a spreadsheet surprise. When validity falls, the resource leaves the available pool — with a record of when and why.
3. Override is a governed exception. If quality authorizes proceeding with a caveat, that becomes a deviation or an explicit permission, attributable, not a verbal “ok” on the radio.
4. Evidence survives inspection. State, validity, and authorization must be reconstructible for that batch at that time — not “in general the plant controls this.”

This is where a resource-readiness module (on NEO, NEO TASK) stops being a task list and becomes part of the execution system of record: validity, state, and blocking of what is nonconforming. The same moment-of-control logic shows up in electronic batch records.

What software does not fix alone

No block saves a dishonest master data set. If cleaning validity was parameterized without a criterion, if CMMS calibration does not talk to the floor, or if the LMS skill is not tied to the EBR step, the system only enforces a fiction faster.

It also does not replace quality’s decision on residual risk. It imposes the moment of the decision. Without an owner of state and without transition rules, “automatic block” becomes noise the operation learns to override.

Questions before calling resources “under control”

• Does the step start query state and validity in real time, or trust the morning checklist?
• When calibration, cleaning, maintenance, or skill expire — does the resource leave availability, or only raise a reminder?
• Do overrides require attributable authorization and stay on the batch record?
• Do person readiness and equipment readiness use the same validity logic, or is one rigid and the other a workaround?
• In inspection, do you show the block that prevented use — or the investigation of use that should never have happened?

That last question usually separates theater from control.

Resource readiness done well is not another list to sign. It is the plant refusing, by design, to start what it cannot defend later. A checklist asks. Control prevents.

In your operation, is “ready” still a mark on paper — or a condition the system can deny?