Unit stance: KEEP, SERVICE, RETIRE, REPLACE, or WATCHING

What a unit's stance is

Every unit in your fleet, every dehumidifier, air mover, air scrubber, moisture meter, or truck mount, gets one of five stances the moment you open its detail card. The stance is Verinode's read on what to do with that specific unit right now, built from its age against an industry-standard expected lifespan, its calibration cycle, its logged condition, whether a newer catalog model exists for its class, and (for rented units) how long it has been on rent. It is not a generic reminder. It is the same call a fractional COO would make looking at that one unit's file.

The five stances are:

• Keep, the unit is healthy, inside its expected working life, and doing its job. Nothing to do but keep logging service events.
• Service, calibration is overdue, condition has slipped, or the unit is closing in on end-of-life but still worth running if you fix what is fixable first.
• Retire, the unit is past its expected lifespan, or it needs repair and is already well into the back half of its life. Plan the exit rather than sink more into it.
• Replace, the unit is aging and the catalog already has a successor model. The market moved on. Swap it, do not renew it.
• Watching, Verinode does not have enough data, usually a purchase date, to stake a stance yet. This is a real, deliberate state, not a bug or a blank screen.

Verinode never decides for you. It surfaces the read, the reasoning behind it, and what is at stake in dollars where that is knowable. You decide whether to service, retire, replace, or keep running the unit.

Where to find it

Open Equipment from the sidebar (/equipment), then click into any unit's card. The stance lives in three places on that unit's detail view, all driven by the same calculation so they never disagree with each other:

1. The stance pill next to the equipment class name at the top of the hero.
2. The subtitle line under the unit's name, a one-line action statement.
3. The agent insight card, the block directly under the hero, which expands the same call into a full headline, evidence, and (when there is one) a dollar figure at stake.

The hero

Stance pill

Next to the unit's equipment class (for example "LGR Dehumidifier"), a small rounded pill reads KEEP, SERVICE, RETIRE, REPLACE, or WATCHING in uppercase. Its color is not decorative, it maps to the brand's signal palette:

• Keep renders in Deere Green, the same accent used for Expand signals elsewhere on the platform.
• Service renders in Hard Hat Yellow, the Maintain signal color.
• Retire renders in Ember Red, the Analyse signal color.
• Replace renders in Copper, the platform's accent color, because it is a "move to the newer thing" call rather than a warning.
• Watching renders muted, because it is a data gap, not a posture.

The unit's overall hero accent (the photo frame, the action button styling) follows the same color, so the whole card reads as one consistent statement about that unit.

Subtitle: the action line

Directly under the unit's name sits a short, concrete instruction, the same one that drives the agent insight card's headline. Examples of what you will actually see, straight from the stance engine:

• Keep: "Keep deploying, this one's earning"
• Service (calibration): "Schedule calibration"
• Service (condition or approaching end-of-life): "Schedule service before the next deployment"
• Service (rented and past the rent-versus-buy break-even window): "Close the rent-vs-buy gap", with a subtitle of "Buy or Return"
• Retire: "Plan the exit on this unit"
• Replace: "Swap to the [catalog model name]", naming the actual successor product
• Watching: "Log a purchase date to unlock a stance"

Hero stats

Four stat tiles sit across the top of the hero, each with a value and, where there is context to compare against, a delta line underneath:

• Age, the unit's age in years to one decimal, with a delta reading "of Ny expected" where N is the industry-standard expected lifespan for that equipment class (for example a portable extractor is benchmarked at 6 years, a truck mount at 8, a thermal imaging camera at 8). The tile tones green when the unit is comfortably inside that window, amber as it closes in on end-of-life, and red once it has run past its expected life. If no purchase date is on file, this reads ", ".
• Calibration, reads OK or Overdue for equipment classes that carry a calibration cycle under IICRC guidance (moisture meters, thermal imaging cameras, hygrometers, psychrometers all recalibrate on a 12-month cycle). The delta line names the cycle length, for example "12mo cycle." Equipment classes with no calibration requirement (air movers, dehumidifiers) show "n/a".
• Same-class units, how many units you own in this same equipment class, including this one. When Verinode has a peer average to compare against, the delta reads "peers: N avg" and tones green if your count is at or above that average, red if you are running fewer units in the class than peers typically carry, a capacity signal worth noticing on a unit that is already trending toward Service or Retire.
• Purchase, the purchase price you logged, with a delta showing how long ago the purchase was ("3mo ago," "2y 4mo ago," or "this month" for a fresh buy). Reads ", " if no price is on file.

Photo and Retire/Restore

The circular logo slot in the hero doubles as the unit's photo. Click it to zoom an existing photo full-screen, or click Add photo (or Change photo) underneath to upload one from your device (10 MB limit). Photos help crews confirm they grabbed the right unit off the shelf, not just the right model number.

The hero's action row carries a Retire button (it reads Restore if the unit is already retired). Clicking it opens a confirmation: retiring a unit marks it retired and drops it off your active equipment list, but its data (age, service history, everything) is preserved and you can restore it later. This is a manual status flip you make, independent of what the stance engine is currently recommending, a unit can carry a Retire stance for months before you actually act on it, and you can also retire a unit the stance engine still calls Keep (for example, if you are downsizing a class).

The agent insight card

Directly under the hero sits the agent insight card, the same second-position surface every entity detail view in Verinode carries. On equipment, this card always has something to say, even when the stance is Keep, because the whole point of the stance engine is that the agent has a live opinion on every unit, not just the ones with an open Finding.

What you see:

• A colored mood badge and an eyebrow label: Recommended (Ember Red, shown for Retire and Replace stances), Watching (Hard Hat Yellow, shown for Service and Watching stances), or On track (Deere Green, shown for Keep).
• A headline, the same action line as the hero subtitle ("Schedule calibration," "Plan the exit on this unit," and so on).
• Evidence, one or two lines explaining the why in plain terms, followed by a second line naming the upside of acting. For example, on a Retire call: "4.2 years old and the catalog already has a successor, you're running last-gen equipment on current-gen jobs" followed by "Move this unit to a secondary route or sell, and deploy the newer model where dollars-per-hour matter most."
• Where a dollar figure is known, an impact chip next to the eyebrow reads "· $X/mo at stake." This only appears when the loss is genuinely dollar-quantified, today that is the rent-versus-buy Service case, where the chip shows the unit's approximate monthly rental rate. Structural calls (aging, calibration overdue, condition poor, Replace) do not carry a fabricated dollar figure, they show the reasoning without a chip instead of forcing a number that is not real.
• If a live linked decision already exists for this unit (see Findings below), the agent insight card shows that decision's own action title and reasoning instead of the raw stance text, so you are never looking at two slightly different versions of the same call.

The action, gain, and cost-of-inaction triad

Every stance carries three linked pieces of language, and you will see all three represented between the hero subtitle and the agent insight evidence:

1. Action, what to actually do (the hero subtitle and insight headline): schedule calibration, schedule service, plan the exit, swap to the named model, keep deploying, or log a purchase date.
2. Cost of inaction, the consequence line explaining what happens if you do not act (the first evidence line): a reliability risk on every job you send an overdue unit out on, a repair bill that is likely to outrun the unit's remaining value, running last-gen equipment on current-gen jobs, or (for Watching) simply that Verinode cannot reason about age, calibration, or rental break-even without a purchase date.
3. Gain, what you get by acting (the second evidence line): a calibrated instrument instead of a callback, cheaper preventive service instead of a mid-job failure, a healthier unit on the route while the new one earns where it matters, or unlocking the stance call entirely by adding a purchase date.

This triad is always present, not just when a decision has been formally opened, because a stance-only Retire or Service call is still real information worth acting on today.

Findings

Scroll down (or tap the Findings tab) to the Findings section for a fuller list of any decisions the agent has actually opened on this unit, separate from the always-on stance triad above it. A Finding is a formal, tracked decision, the stance is the agent's live opinion; a Finding is what happens when that opinion has been escalated into something with its own action plan and (often) a dollar figure attached.

What you see, when decisions exist. A header line: "N decision(s) used this record as evidence." Then one row per decision, each showing:

• A status label: Recommended, Watching, or For review.
• The decision's action title, in plain language.
• A short compressed reason underneath, when one is logged.
• If the decision carries a monthly loss figure, it appears on the right as a bold dollar amount with "/mo."

What you see while loading. "Gathering decisions for this unit…"

Empty state, verbatim. "No linked decisions on this unit right now. If a signal trips (aging, calibration, rental break-even), it'll land here alongside the action plan."

An empty Findings section is normal and expected for most units most of the time. The stance triad above it is doing the day-to-day work; Findings is where a signal that has crossed into something worth formally tracking shows up alongside its action plan.

How to use it

1. 1Open Equipment from the sidebar and scan the stance pills across your fleet cards, Retire and Replace pills are your priority queue.
2. 2Click into a unit carrying Retire, Replace, or Service and read the hero subtitle plus the agent insight evidence, that is the full reasoning in two sentences.
3. 3Check the hero stats: Age against expected lifespan, Calibration status, and Same-class units against the peer average, to see whether the stance is an isolated call or part of a fleet-wide pattern (for example, several units in the same class all closing in on end-of-life together).
4. 4If the stance is Replace, note the named successor model in the action line, that is a real catalog product you can look up under the unit's Alternatives tab.
5. 5Check Findings for a formally tracked decision on the unit before you act, if the agent has already opened one, work from its action plan rather than duplicating the stance triad's reasoning.
6. 6Decide: service, retire, replace, or keep running it. If you retire it, use the Retire button in the hero, the unit's history stays intact and you can restore it later.

Data sources

The stance engine reads your own fleet data (purchase date, condition, ownership type, calibration and maintenance dates) against reference tables Verinode maintains: an industry-standard expected lifespan per equipment class, an IICRC-aligned calibration interval for instrument classes, and the equipment catalog's newer-model linkage. Peer averages shown in the "Same-class units" delta come from anonymized fleet data across the operator network. As an independent data trust, Verinode never sells this data, or your fleet's data, to carriers or vendors; benchmarks like the peer average exist only to give you a read on your own fleet.

Related reading

• How benchmarks work
• Reading a benchmark
• The decision workspace
• Understanding your margin