Assumption Is Not Evidence

Facts come before assumptions. Clearance, authorization, and learning require a recorded check or a verified outcome. A confident guess is not that record. On the Decision Case, an assumption can stay visible if it is labeled as an assumption. It does not become observed fact because the room is sure.

The handover says the gearbox was fine. The crew says they always stretch this preventive-maintenance interval. The model ranks the asset for deferral, and the tile looks confident. The meeting treats those sentences as the evidence and moves the case. They are not the evidence. The handover is a story. The interval is a habit. The rank is an output that has not been tied to the work history, the condition record, or the last check that produced it. A later reader cannot inspect a fact the case never held.

This essay is the companion to Unknown Is Not Clear, Blank Is Not Zero, and Silence Is Not Clearance. Unknown Is Not Clear keeps missing evidence recorded as unknown, not as clear or ready. Blank Is Not Zero keeps an empty field from being stored as a measured zero. Silence Is Not Clearance keeps no reply, no alert, and an empty inbox from being read as approval. This essay is the statement people put in the gap. An assumption does not fill unknown and make it clear. It does not turn a blank into a measurement. It does not turn quiet into a named decision.

Orville Davis states the Decision Case order in Field Manual v0: Question, Evidence, Recommendation, Human decision, Action, Verification, and Learning. The manuals index lives at /manuals. Evidence is step 02. The Evidence chapter keeps a visible split between observed fact, assumption, and what is still missing. The recommendation waits on that record. A proposal whose only basis is an unlabeled assumption has no basis a later reader can inspect. A proposal with no basis is not ready for a human decision. Honesty boundaries sit beside the spine: recommend is not authorize, evidence is required, and Stage-1 does not treat a model line or a crew habit as a live fact.

An assumption is not a recorded fact

A recorded fact is something a later reader can inspect on the case: approved knowledge, asset context, work history, condition information, or a check that was written down. An assumption is a statement about the asset or the work that does not have that record behind it. The Decision Case can hold the statement. The operator labels it as an assumption. The label is the mechanism. It keeps the statement from being stored in the observed-fact column.

A belief is someone being sure. The alarm is a nuisance. The bearing will last the run. The last repair fixed the cause. Sure is a state of the person. It is not a recorded observation of the condition. The belief can be written on the case so the next reader sees what was believed. Writing it down does not make the condition what the person believed. Condition information is the reading, the inspection, or the work record. The belief is the other column.

A habit is a repeated practice. The crew changes this class of component on a fixed interval, or they defer this class of alarm. A habit explains why a recommendation was drafted. It is not the failure record, the meter reading, or the last verified outcome. A CMMS row can show a last-done date if that date was entered against a closed work order. The sentence "we always do it at this interval" does not replace that date, and it does not show that the interval is still fit for this asset. The row is the record of what was entered. The habit is not a second measurement hiding inside the row.

A prior shift's story is the handover. The pump was fine. The last job fixed it. The part is on the way. A story is what someone said at the relief. Work history on the case is the closed record of what was done and what was checked. If the prior case has no verification, the handover does not supply the check. The operator does not promote the spoken relief into observed fact. The story can sit on the case as an assumption, attributed to the shift that said it. Attribution is not verification.

A tribal "we always do it this way" is the practice of the crew, the site, or the trade. It can be true as a description of what people usually do. It is not evidence that the practice is the right next action for this asset, this question, or this shift. The Decision Case does not treat the usual practice as clearance. Usual is not a recorded check. A later reader who sees only the phrase cannot reconstruct which asset configuration, which failure, or which maintenance record justified the work.

An unverified model output is a rank, a health score, a suggested deferral, or a stretched preventive-maintenance interval that the case cannot reconstruct. Evidence Lineage Is Not Optional already requires that trail: which asset configuration, which failure and maintenance records, and which human judgment produced the recommendation. An output without the trail is an assumption wearing a number. The dashboard can be confident. Confidence Is Not Evidence is that sentence at full length: a strong gut feel, a high model score, years of tribal "we've always known," or a manager's certainty is still not a recorded check or a verified outcome. Confidence is not the lineage. Stage-1 evidence is the record held on the case. A live connector that pulls historian or control-system tags sits outside this edition. A score that arrived without those records is not a tag, and it is not observed fact.

An assumption can stay on the case, labeled. The label does not promote it to a recorded fact.

Facts before assumptions

Clearance is a recorded basis the next step can stand on. The evidence is present and fit, a named person has decided, and, where an action was authorized, the verification is on the case. An assumption is not that basis. A confident guess does not clear missing evidence, does not measure an empty field, and does not approve a recommendation nobody answered.

Unknown Is Not Clear records what is missing as unknown, and it blocks evidence that is present and bad or not ready. Filling the unknown line with a belief, a handover, or a model rank does not change the state to clear. The missing record is still missing. The assumption is an additional statement. It belongs in the assumption column, or it stays off the case. It does not overwrite unknown.

Blank Is Not Zero keeps an empty field, a missing reading, or an unset value from being stored as a measured zero. Zero is a verified measurement of none. An assumption that there was no downtime, no cost, or no delay is not that measurement. Typing the assumption into the blank collects a guess. It does not collect the value. Recovery economic figures that were left blank stay blank. They are not inferred from what the last shift thought the job should have cost.

Silence Is Not Clearance keeps silence, no reply, no alert, and an empty inbox from being read as authorization. A habit that "nobody objected last time, so this is approved" is an assumption about a missing response. The missing response is still not approved. The assumption does not write accept, reject, escalate, or return on the decision line. A quiet dashboard is still not controlled.

Authorization is the named act. A person accepts, rejects, escalates, or returns the recommendation, and that act is recorded beside the proposal. Human Decision Is Not Optional says the case stays a proposal until the act is on the record. Recommend Is Not Authorize says a drafted next action is not that act. A model line that says defer, and a crew habit that says defer, are both still proposals. Neither one accepts the operational consequence for the person who can accept it. The system does not treat the assumption as the acceptance.

Learning is what a later question is allowed to inherit. The closed case shows what was believed, what was authorized, and what was checked. Verification Is Not Optional says authorization records who decided, and the case stays open until the check is written down. Learning Requires a Verified Outcome says the later question inherits that verified record, not a hope. What was believed may include a labeled assumption. Learning does not inherit the assumption as if the check had confirmed it. A later crew that repeats the story because "it worked last time," without the verification on the prior case, is inheriting a story. The Learning chapter publishes the closed record. It does not publish a guess as a result.

Clearance, authorization, and learning stand on a recorded check or a verified outcome. A confident guess does not stand in for either record.

Where an assumption sits on the Decision Case

The order is Question, Evidence, Recommendation, Human decision, Action, Verification, and Learning. An assumption does not complete a step by being confident. It stays labeled, or it stays off the record.

Question is step 01. An assumption is not the question. "We already know what to do" does not open the case, and it does not close a question that was never written down. Question Is Not Decision says recording a question is not the decision. Skipping the question because the crew is sure is not a decision either. The Question chapter still requires the question on the case before evidence or recommendation.

Evidence is step 02. This is the step the assumption tries to skip. Observed fact, assumption, and what is still missing stay in separate columns. Merging the assumption into observed fact is the failure. A CMMS comment, a failure code, and a last-done date are records of what was entered. A comment that repeats the tribal practice is not condition evidence. Work that was closed with a recorded check can be observed fact about that past work. The story told about that work, without the check, stays an assumption. The Evidence chapter keeps the split visible so the recommendation cannot hide which column it used.

Recommendation is step 03. The proposal states a next action, the evidence it uses, and the uncertainty it leaves in view. If the basis is an assumption, the recommendation says so. Drafting the next action does not authorize the work, and it does not convert the assumption into fact. A rank the model produced is part of the proposal, or it is part of the uncertainty. It is not a silent upgrade into the evidence column. The Recommendation chapter leaves authority where it already sits.

Human decision is step 04. A named person accepts, rejects, escalates, or returns. Accepting a recommendation that rests on a labeled assumption is a decision about that proposal. It does not rewrite the assumption as observed fact. The person who decides can see the split. The system does not hide the assumption so the decision looks better supported than the record is. The Human decision chapter keeps who decided visible beside the recommendation. An assumption about what that person would have wanted has no name on the decision line.

Action is step 05. Action records the work the human decision authorized. It does not turn the assumption into a completed repair, a written work order, or an isolation. Plant execute stays outside Stage-1. A confident deferral on a screen does not dispatch the work, and it does not mean the work was done. The Action chapter keeps that work with the systems of record the customer already runs, and with the people who already hold operating authority. Action Is Not Execution keeps the same boundary. An assumption does not cross it.

Verification is step 06. The check records whether the authorized action did what it was supposed to do, and which evidence would show that. An assumption that it worked is not the check. If the expected signal was not available, that unavailability is written down. It is not filled with a guess of success. The case stays open until the check is on the case. The Verification chapter is that recorded check. Stage-1 does not treat a live tag pull, or a model score standing in for one, as the proof.

Learning is step 07. A later question can see the labeled assumption and the check that followed it, or the absence of the check. It cannot treat the assumption as the outcome. There is no verified outcome in a handover that was never checked. The Learning chapter publishes that rule.

Each step needs its own record. An assumption does not complete the step it is standing in for.

The neighboring boundaries still hold

This boundary is not the neighboring ones. Each of those essays still holds. None of them promotes an assumption to a recorded fact.

Verification Is Not Optional is the check after a named human has authorized an action. An assumption is earlier and different when it is used as evidence, and it is still not the check when someone uses it to close the case. The operator writes the observation. A belief that the work succeeded leaves the verification field empty. Empty stays empty until the check is recorded.

Evidence Lineage Is Not Optional is the trail behind a recommendation. A model output, a bad-actor tag, or a stretched interval that cannot show its inputs is a claim with the trail missing. This essay is the content of that claim when the content is a belief, a habit, or a story. Lineage does not appear because the output is numeric. The records are either on the case or they are not. If they are not, the output stays an assumption.

Honesty Boundary Is Not Optional says Sync states what is simulated, seeded, or out of scope. A simulated or seeded score is a practice record. It is not a live plant fact, and it is not a substitute for the assumption the operator still has to label. Naming the limit does not promote the practice record into observed fact. The Honesty boundaries chapter keeps Stage-1 from claiming a live result the edition does not hold.

Where the public statement lives

Field Manual v0 is the public contents of the Decision Case. Start at the manuals index or open Sync Field Manual directly. Evidence is step 02. Observed fact, assumption, and what is still missing stay distinct. A proposal with no basis is not ready for the human decision that follows. Verification is step 06. A guess that the action worked is not the check. Learning is step 07. A later case inherits the verified record, not the assumption. Honesty boundaries keep the edition from treating a model line or a crew habit as a live fact. Later editions can deepen a chapter. The spine stays in this order.

What this article is not claiming

This is an essay about the record, not a customer case study. It names no plant, states no savings figure, and claims no prevented failure. It states no OEM limit and no operating threshold. It does not claim that Sync executes plant work, writes a work order, isolates equipment, or controls a plant. It does not claim that an assumption is evidence, that a belief or a habit is a recorded fact, or that an unverified model output is a verified outcome. It does not claim that self-guided onboarding is a live product path.

Stage-1 readiness means a signed-in user can complete the Decision Case — question, evidence, recommendation, human decision, action, verification, and learning — and Field Manual v0 describes that journey. This edition does not describe plant execute, a live connector tag pull, SMTP invite delivery, or automatic revocation of access on expiry as live. It does not describe Sync writing work orders, isolating equipment, or controlling the plant. Simulated or seeded telemetry and assets are practice records. They are not live plant results, and a model line on a practice screen is not observed fact. Self-guided onboarding is not claimed as a live product path.

The series continues with Confidence Is Not Evidence, on why a strong gut feel, a high model score, years of tribal "we've always known," or a manager's certainty is still not a recorded check or a verified outcome. Companion reading: Unknown Is Not Clear on why missing evidence is recorded as unknown, not as clear or ready, Blank Is Not Zero on why an empty field is not a measured zero, Silence Is Not Clearance on why silence, no reply, no alert, and an empty inbox are not authorization, Verification Is Not Optional on why the case stays open until the check is recorded, and Evidence Lineage Is Not Optional on why a recommendation you cannot reconstruct is not an industrial decision. A Reliability Assessment asks whether the records can support a conclusion. A Strategic Pilot is a governed proof around one operating decision. The Evidence chapter is the public statement of the split between observed fact, assumption, and what is still missing.