Build the case on numbers your finance team already defends.
Five operating assumptions and one investment figure, all of them yours. The model
returns an annual value, a break-even month, and a link your team can check line by
line. Nothing is calculated until you enter a number, and nothing you enter leaves
your browser.
Maintenance sets the operating assumptions. Operations validates the impact. Finance
approves the inputs and owns the decision.
01Downtime value = events x avoidable hours x cost per hour
02Labor value = events x coordination hours x loaded rate
03Total modeled value = downtime value + labor value
04Break-even months = 12 x investment / total modeled value
05Modeled ROI = (total modeled value - investment) / investment
That is the entire model. There is no hidden multiplier, no benchmark, and no
assumption we did not ask you for.
Finance controls every input
We supply no cost per downtime hour, no loaded labor rate, and no price. A number you cannot defend in a budget review has no business in the model.
The arithmetic stays visible
Two multiplications, one division, one break-even. Read them, argue with them, and reproduce the whole model in your own spreadsheet in five minutes.
Nothing is claimed on your behalf
Every input and output remains visible as your arithmetic on your assumptions. Our measured results are stated separately, further down.
EQUA AIMMS customer-specific value model
Modeled operating value and break-even
Prepared on the date shown by your browser. Model published at
equa.work/roi. Every input below was supplied by the reader, not by EQUA AI.
Your assumptions
Start with finance's assumptions, not ours.
Enter the operating assumptions your team can defend. The formulas stay visible, finance
controls every input, and no value appears until you supply the numbers.
This model runs entirely in your browser and needs JavaScript to calculate. The five
formulas above are the complete model, so you can reproduce it in a spreadsheet without
running anything from this site.
Modeled annual value from your inputs
Total modeled value per year
$0
Break-even at these assumptions
Enter investment
Enter your assumptions to produce a result. Nothing is calculated until you do.
Downtime avoided
events x avoidable hours x cost per hour
$0
Labor capacity recovered
events x coordination hours x loaded rate
$0
Modeled ROI
(total modeled value - investment) / investment
Enter investment
This is your arithmetic on your assumptions, not a measurement, forecast, guarantee, or
AIMMS result. A measured first deployment replaces the assumptions with your operation's
baseline and measured change. Finance controls every input.
Optional refinement
Which delay hours do you believe are removable?
The six inputs above are enough on their own. This layer exists only to help you
defend the avoidable-hours figure. Nine of the twelve delay categories are digital
coordination AIMMS can move. Three are physical work that stays with people, so the
model never counts them. Change nothing and the pre-selection stands.
Hatched figures show the modelled structure. Delivery counts only the
ordering and coordination time before a shipment moves, never transit time.
Selected scope suggests 5.0 avoidable hours per event.
Inputs supplied by the reader
Model published at equa.work/roi. Every figure above uses assumptions supplied by the
reader; EQUA AI supplied none of the inputs. A first deployment replaces those assumptions
with a real baseline and a measured change.
Send the model, not a screenshot.
The link below reproduces this exact model: every input, and the delay categories you
selected. Anyone who opens it sees the same arithmetic and can change any assumption.
Privacy: the link carries your assumptions after the #, which browsers
never send to a server. They stay in the recipient's browser exactly as they stayed in
yours: not in our logs, not in any analytics event, and not in any referrer this page
hands onward. Treat the link itself like any other internal financial document, because
anyone you send it to can read the numbers in it.
How to defend this in a budget review
A conservative model survives the meeting. An optimistic one does not.
01
Two levers only
The model counts downtime avoided and coordination capacity recovered. It deliberately ignores procurement leakage, expert load, record quality, and retained knowledge, because those are easier to claim than to defend. If two levers do not carry the case, the case is not ready.
02
Your cost of adoption belongs in the investment figure
Put integration work, data cleanup, security review, training, and internal time into the investment input. A break-even that excludes your own effort is not a break-even.
03
The control boundary does not move
The model assumes no change to who operates the plant. AIMMS moves permitted digital work. It does not write to PLC, DCS, or SCADA, and it does not execute setpoints, restarts, interlocks, or actuator commands. SCADA and historian context stays read-only.
04
Measurement replaces assumptions
The only way to replace these assumptions is to measure them. A first deployment reconstructs your baseline, runs the workflow for a focused operating period, and reports the change against the number you entered here.
Measured in one anonymized industrial production deployment. They are not a result in your sector, a forecast, or a universal guarantee. Results vary by workflow, asset mix, data quality, operating environment, and customer controls.
Measured first deployment
From baseline to rollout decision.
01
Baseline
Reconstruct recent repairs and quantify where time is spent.
02
Onboard
Target a customizable two-week setup for one consequential workflow, subject to data, security, and user readiness.
03
Validate
Run historical cases and confirm evidence, parts, workflow, and control quality.
04
Operate
Use AIMMS on the approved live workflow for a focused 30-day operating period after readiness.
05
Measure
Compare downtime, ready-work time, expert load, supplier cycle, approvals, and closeout.
06
Decide
Produce a finance-ready rollout case, control model, and expansion scope.
What the waiting is made of
The alarm is not the repair.
Detection is instant. What follows is not. People search for context, verify safety,
identify the right part, coordinate suppliers, find authority, and reconstruct the
record. The repair is the short part. Draw the same repair twice and you can see
where the days actually go.
AThe repair todayWaiting first. Work last.
EvidenceReadinessPartsAuthorityApprovalRepair
Fault detected
The repair starts here instead
BThe same repair, with AIMMSThe same five stretches, compressed.
EvidenceReadinessPartsAuthorityApprovalRepairTime returned to the operation
Fault detected
Waiting. Evidence, readiness, parts, authority, approval. Hatched because it is drawn structure, not a measured quantity.
The repair. Same width in both records: the same physical work, by the same qualified people.
Verified return. The closing edge of the repair block.
Time returned. Ground the asset spends running instead of waiting.
Nothing about the repair changes.
Both records draw the repair block at one shared width, so the work cannot appear to
get shorter and cannot appear to change hands. Physical work and safety-critical
authority stay with qualified people. The only thing this figure moves is when the
work is able to start. It contains no numbers: it is structure, not measurement.
Physical judgment stays in qualified hands.
Four things go missing between the alarm and the fix.
These are the gaps the drawing above is made of. Each one is a place the work is
ready and the decision is not.
01EvidenceSignals, manuals, history and observations sit in different places.The technician proves what the asset already told someone else.
02ReadinessSafety, people, parts, suppliers and access do not become ready together.Four things are ready and the fifth decides the start time.
03AuthorityThe next action waits because the right approval and boundary are unclear.Nobody is refusing. The decision simply has no owner in front of it.
04MemoryThe result closes in fragments, so the next fault starts with rediscovery.The same fault costs the same hours the second time it happens.
Your team is not slow. The work around the repair is. Put in your last critical
repair and see the split. We do not project a result onto your numbers.
Drag both controls. Use your own last repair.
This section is a calculator. Without JavaScript the controls cannot compute, so the
figures beside this paragraph are a worked example on a 38-hour repair rather than
anything you entered. The point it makes does not depend on the arithmetic: on a
critical repair, most of the elapsed time is spent waiting on coordination rather than
on the asset.
The ledger uses only the two values you enter and applies no assumed AIMMS effect.
Your repair 38 h
Hands on the asset
Waiting on coordination
32 of those hours were waiting. 84% of that repair was coordination, not repair. Your crew was ready for
6 of those 38 hours.
What we measured
In one anonymized industrial production deployment, MTTR fell
14.2 h9.7 h,
a 31.7% reduction.
Measured in one anonymized industrial production deployment. They are not a result in your sector, a forecast, or a universal guarantee. Results vary by workflow, asset mix, data quality, operating environment, and customer controls.
Do not treat MTTR as one number. Find the hours that are actually removable.
The assessment classifies every elapsed hour of your recent repairs into twelve delay
categories. The baseline shows exactly which delays AIMMS can remove, which must remain,
and where the first deployment should focus.
01 Asset access
02 Diagnosis
03 Information search
04 Expert assistance
05 Part identification
06 Inventory confirmation
07 Supplier response
08 Approval
09 Delivery
10 Physical repair
11 Testing and restart
12 Documentation and closeout
Digital coordination AIMMS can remove or compressPhysical work that stays with people
The line between the two halves of this page
Your model ends here. Nothing below it was used to produce anything above it.
Every figure above is your arithmetic on assumptions you supplied, with no AIMMS effect
applied. Every figure below is measured, in one anonymized industrial production
deployment, and it is not a forecast of your result.
Not a model input
Separate from your model: what we measured.
Everything above is your arithmetic on your assumptions. The figures below are ours. They
are measured, they are reported at their real scope, and they are not fed into the model,
not used as a default, and not projected onto your operation. Read them as evidence that
the workflow moves, then decide for yourself what it would move at your site.
Measured production evidence
One anonymized industrial production deployment
Live since
Jan 2026
Critical assets
10
Active production users
5
Operating cadence
Daily
Workflow control
Approval-gated
lower MTTR
31.7%
14.2 h to 9.7 h
Measured on the deployed repair workflow.
faster quote cycles
80.4%
4.6 d to 0.9 d
Measured on the deployed quoting workflow.
faster quote-to-order
96%
3 d to less than 1 h
Measured on the deployed quote-to-order workflow.
lower measured safety-risk exposure
50%
Approval-gated workflows
Measured on workflows with defined approval gates.
Measured in one anonymized industrial production deployment. They are not a result in your sector, a forecast, or a universal guarantee. Results vary by workflow, asset mix, data quality, operating environment, and customer controls.
Where value comes from
The operating levers behind the ROI.
Six levers move when the digital work moves. The value model on the ROI page counts only
the first two, because they are the two a finance approver can defend without argument.
Counted in the model
Downtime avoided
The operation changes Approved, executable work starts sooner after the fault.
Which removes Hours of unavailability, at a cost per hour finance approves.
Counted in the model
Labor capacity recovered
The operation changes Coordination, search, retyping and follow-up stop being the technician’s job.
Which removes Hours at a loaded rate finance approves.
Deliberately not counted
Procurement leakage reduced
The operation changes The part is confirmed against the asset before anything is ordered.
Which removes Expedites, wrong parts, emergency buys, duplicate orders.
Deliberately not counted
Expert load reduced
The operation changes The context a senior engineer would have rebuilt is already assembled.
Which removes Escalations, repeat troubleshooting, after-hours dependency.
Deliberately not counted
Record quality improved
The operation changes Closeout is captured while the work happens rather than from memory.
Which removes Incomplete records, stale inventory, unrecorded substitutions.
Deliberately not counted
Knowledge retained
The operation changes What the crew learned stays attached to the asset.
Which removes The cost of rediscovering the same fault next time.
Four of these six are real and none of them is in the arithmetic. A model that counts
everything it can imagine is not a stronger case, it is an easier one to reject.
One decision chain
Operating proof that maintenance, operations, and finance can use.
Maintenance and reliability
Own the operating problem
Choose the recurring failure, validate the current delay, and judge whether the workflow becomes easier to execute.
Operations and the COO
Own service and operating impact
Confirm the control boundary, return-to-service evidence, and effect on continuity and team capacity.
Finance and the CFO
Own the investment decision
Approve the economic inputs and compare measured operating value with the cost and risk of rollout.
Bring the model. Leave with a baseline that replaces it.
In 20 minutes we will map the current delay, the customer control boundary, and a first-deployment measurement plan that reports against the numbers you just entered.
One operating problem. One focused working session. No obligation.
EQUA uses essential cookies and limited analytics to understand site performance. You can accept
analytics or continue with essential cookies only. See our Privacy Policy.