When EQUA AIMMS Is the Wrong Purchase
Seven situations in which EQUA AIMMS is the wrong thing to buy, what to buy instead, and what EQUA AI has not published that a buyer should weigh.
Working on a live operating problem? Book Your 20-Minute Assessment
Seven situations make EQUA AIMMS the wrong purchase: no read access to SCADA, a historian or sensor data; no maintenance system worth writing back to; no defined approval structure; a real need for condition monitoring; a need for software to act on the control system; one or two critical assets run by one person; and a procurement that requires a named reference customer, which EQUA AI does not have.
Key takeaways
- Two of the seven are absolute. No permission level in EQUA AIMMS will change a setpoint or restart a pump. And vibration, thermography, oil and motor current analysis are a separate discipline with their own instrumentation and specialists.
- AIMMS does not replace the CMMS, and it is a poor first maintenance system. The CMMS, EAM or ERP stays the system of record. If there is no record system, buy one first.
- Autonomy is customer-defined, which means it is customer-specified. If nobody can state who may approve what, and up to what value, there is nothing to configure and the gates become decoration.
- EQUA AI has one anonymized production deployment, no named customer, no published case study and no product screenshots. Some procurement processes cannot accept that. Weigh it as a fact about the vendor in front of you.
- The measurement method behind EQUA’s four published figures has not been published in full. The deployment scope is stated on the value model page, and the business-hours clock behind the quote-to-order figure is stated in this site’s benchmark provenance ledger. The rest of the protocol is not. That is a current gap, not a forthcoming document.
- Coordination overhead has to exist in volume before software that reduces it pays for itself. One operator and two critical assets does not clear that bar.
What are the seven disqualifiers?
| Situation | Why it disqualifies | What to buy or do instead |
|---|---|---|
| No read access to SCADA, a historian or sensors | Detection has nothing to read | A historian, telemetry or an integration project |
| No CMMS or EAM, or one nobody uses | No system of record to write back to | A CMMS or EAM, adopted first |
| No defined approval structure | Nothing to configure; gates become theatre | Write the delegation of authority |
| The real problem is degradation detection | Different discipline, instruments and specialists | A condition monitoring programme |
| You need a control-system action | No write path to a PLC, DCS or SCADA system, ever | Your control system integrator |
| One or two critical assets, one person | Too little coordination overhead to be worth changing | Nothing. Revisit as the operation grows |
| Procurement requires evidence EQUA lacks | No named sector reference or published certification exists | A vendor with the required track record |
Can EQUA AIMMS read your operational data?
AIMMS connects read-only to SCADA, historians and IoT sensor sources. It detects anomalies and deviations in that data, can raise the fault itself and alert the responsible people, and surfaces slower drifts as recommended scheduled maintenance. All of that starts from a read.
Take the read away and half the product has nothing to work from. No historian. A policy that will not permit a mediated read across. An integration project sitting unfunded in next year’s budget. Any of those, and detection is blind and the coordination half starts every job blind with it. Black and Veatch’s press release for its 2026 Water Report, a survey of more than 600 US water industry stakeholders, states that “seven in 10 (70%) say they collect sufficient data, but only 19% say they leverage it effectively”. If you are in the minority that does not collect sufficient data, the sequence is data first.
Is there a maintenance system worth writing back to?
AIMMS does not replace the CMMS. The CMMS, EAM or ERP stays the system of record, and AIMMS writes back into it under the permissions and approval rules you set.
Two versions of this fail. The first is having no maintenance system at all, so AIMMS has nowhere to land its output. The second is more common and harder to say out loud in a procurement meeting: the system exists, it was configured once, and the crew works from a whiteboard and their phones. Writing into a system nobody reads changes nothing.
AIMMS is also a poor first maintenance system. Buying a system of record is a decision about asset registers, work order types, cost coding and whether the crew will actually use it, and it deserves to be made on its own terms against products built for it. Do that first.
Can anyone say who approves what?
Autonomy in AIMMS is customer-defined. Roles, permissions, spend limits, safety policies and approval gates decide what the system may do on its own and what it must route to a person. That is a configuration input, and somebody has to supply it.
An organisation that cannot answer “who may approve a $4,000 emergency part at 02:00, and who if that person is unreachable” cannot supply it. The gates then get set to whatever came out of the implementation call. That answer turns out to be wrong the first time it matters, and the record shows an approval nobody with authority actually gave.
Written policy is often thinner than a procurement conversation assumes. The American Water Works Association’s 2026 State of the Water Industry report, published 30 April 2026, asked about overtime policies and procedures. Of the 770 utility respondents to that item (Figure 16), 13.0% reported comprehensive overtime management, 46.5% basic approval procedures and 32.7% no formal policies or procedures. Read it for exactly what it is. That question is about overtime, not about delegation of authority generally. And the base is a screened subgroup rather than a low response rate: the report states that respondents skipped questions or were not shown certain questions, and it counts 1,273 utility respondents among the 2,171 participants. It is still a reason to open your own written rules before assuming they can be encoded.
Is the real problem degradation detection?
If the assets are not instrumented, or what you actually need is to know a bearing is failing before it fails, buy a condition monitoring programme. Vibration analysis, infrared thermography, oil and wear debris analysis and motor current signature analysis are a separate discipline, with their own instruments, sampling regimes and trained analysts. AIMMS is not a condition monitoring programme and does not substitute for one.
The US Department of Energy’s Federal Energy Management Program guide is blunt about what that discipline costs. Section 6.1 of its chapter on predictive maintenance technologies states that most industry experts, and most reputable equipment vendors, would agree that this equipment “should not be purchased for in-house use if there is not a serious commitment to proper implementation, operator training, and equipment monitoring and repair”. A site that cannot make that commitment, the guide says, is well advised to seek other methods of programme implementation, and it names contracting the services to an outside vendor as a preferable option. That is the decision in front of you, and it is not the one AIMMS is built for.
Detection in AIMMS reads the data you already have: process values, alarms and events, historian trends. It does not put an accelerometer on a pump. A programme run to a recognised standard, ISO 17359:2018, Condition monitoring and diagnostics of machines: General guidelines, is a separate purchase on a separate budget line.
Do you need software to act on the control system?
If you want software to change a setpoint, restart equipment, operate an interlock or command an actuator, AIMMS will not do it. Not at a higher permission level. Not with a signed approval attached. Not in an emergency, and not after configuration. There is no write path to a PLC, DCS or SCADA system in the product.
Detection is a read. Control is a write. The boundary is on the write.
This one disqualifies more cleanly than the rest because no amount of scoping moves it. If closed-loop control is the requirement, that is a control system project with your integrator, inside your own safety case, and it should be procured as one.
Is the operation too small for this to matter?
A site with one or two critical assets and one person who knows both does not have a coordination problem in the sense this product addresses. The parts are in one cupboard. The approver is the person who found the fault. The supplier is a number they have memorised. Software that shortens handoffs between roles has little to work on when there is one role.
There is no citation for this. It is a judgement about where the overhead starts to exist in enough volume to be worth changing anything, and the honest version is that a small operation should spend the money on instrumentation, a spare, or a second trained person first.
Does your procurement require evidence EQUA does not have?
EQUA AI has no named reference customer, no published case study, no product screenshots and no published security certification. If your process requires a reference in your sector, a certification report, or a multi-year corporate track record as a pass or fail condition, EQUA cannot meet it today. There is no longer document that changes that.
Federal negotiated procurement handles the situation explicitly. FAR 15.305(a)(2)(iv) provides that “in the case of an offeror without a record of relevant past performance or for whom information on past performance is not available, the offeror may not be evaluated favorably or unfavorably on past performance”. That rule governs federal negotiated acquisitions. State and local solicitations set their own rules, and many treat references as scored or as a pass or fail condition rather than as neutral. If yours does, the correct outcome is that EQUA does not win. Hold the line. A buyer who writes an exception for a vendor has just moved the vendor’s risk onto their own name.
What has EQUA not published?
Two things, stated as current facts rather than as forthcoming work.
The measurement method behind the four figures, in full. EQUA publishes four measured figures, including 31.7% lower mean time to repair, from 14.2 hours to 9.7 hours, measured on the deployed repair workflow in one anonymized industrial production deployment: live since January 2026, 10 critical assets, 5 active production users, a daily cadence, approval-gated workflows. The scope is published on the value model page, and the clock behind the quote-to-order figure, a business-hours clock, is published in this site’s benchmark provenance ledger. What is not published is the protocol: how each clock was started and stopped, which work orders were in and out of the sample, how the prior period was constructed, and who verified it. Mean time to repair is time to repair, not total downtime, and must never be read as a downtime reduction. Until the protocol is published, give the figures little weight, as you should any vendor figure you cannot audit.
A named customer. The one production deployment is anonymized. No reference call. No site visit. No peer in your sector who will tell you what they would specify differently the second time. For some buyers that is disqualifying on its own, and they are right to treat it that way.
The check that settles it in one question
Scroll sideways to see the whole drawing.
Figure 1. The check that settles it in one question. A decision tree triggered by a utility or plant considering this category of software. The question at the top asks whether operational data can be read and whether there is a maintenance system worth writing back to. The fast branch, taken when both are true, runs: name who approves what and up to what value, pick one asset class and one failure mode, agree what will be measured before anything is installed, and run it against that. The slow branch, taken when either is false, is hatched and runs: fix the read or the system of record first, treat instrumentation coverage as a separate purchase, and come back when there is somewhere for the work to land.
Who is this the right purchase for?
AIMMS fits an operation that already has readable operational data, a maintenance system of record people actually use, and someone who can write down who may approve what. In that setting it works the span between the fault and the closed work order: reading SCADA, historian and sensor sources to detect anomalies and raise the fault, assembling the evidence with each item traced to its source, checking parts against the installed configuration, moving suppliers and quotes, routing approvals to whoever holds the authority, and writing back into the systems you already run.
It suits organisations with enough assets, people and handoffs that the delay between roles is a measurable cost. It does not suit anything on the list above. Those seven are not objections to be handled on a call. They are reasons to buy something else, or nothing.
Who wrote this
EQUA AI builds EQUA AIMMS. It reads SCADA, historians and IoT sensor sources, raises the fault, and then does the digital work between that alarm and a closed work order: the evidence, the parts, the suppliers, the approvals, the record.
Where the conditions above are met, it takes the coordination off your crew and leaves them the repair. Where they are not, it will not help you, and the list above is the one we use to say so on the first call rather than in month ten.
Sources
- Black and Veatch, Evolving water challenges drive innovation: 15th annual Black & Veatch 2026 Water Report highlights path forward, 9 June 2026. Survey of more than 600 US water industry stakeholders. Quoted from the release: “Seven in 10 (70%) say they collect sufficient data, but only 19% say they leverage it effectively.” Note that other figures in the same release are narrower than their headline form; only the two quoted here are used.
- American Water Works Association, 2026 State of the Water Industry Report, published 30 April 2026, survey fielded 21 September to 31 October 2025, total n=2,171. Figure 16, overtime policies and procedures, n=770 utility respondents: 13.0% comprehensive overtime management, 46.5% basic approval procedures, 32.7% no formal policies or procedures. The report counts 1,273 utility respondents among the 2,171 participants and states that respondents skipped questions or were not shown certain questions, so a base of 770 is a screened subgroup and not a response rate. The full report sits behind a request form and this exhibit is not verifiable from the free executive summary alone. Report page and request form: awwa.org/state-of-the-water-industry. Announcement: AWWA State of the Water Industry Report underscores infrastructure, funding challenges, 30 April 2026.
- G. P. Sullivan, R. Pugh, A. P. Melendez and W. D. Hunt, Pacific Northwest National Laboratory for the US Department of Energy Federal Energy Management Program, Operations and Maintenance Best Practices: A Guide to Achieving Operational Efficiency, Release 3.0, PNNL-19634, August 2010. Chapter 6, Predictive Maintenance Technologies, section 6.1, page 6.1.
- US General Services Administration, Department of Defense and NASA, Federal Acquisition Regulation 15.305, Proposal evaluation, paragraph (a)(2)(iv). Applies to federal negotiated procurement; state and local rules differ.
- International Organization for Standardization, ISO 17359:2018, Condition monitoring and diagnostics of machines: General guidelines, third edition, January 2018. Cited by number and published title only; the standard is paywalled, is not quoted here, and nothing is attributed to its contents.
- EQUA AI deployment facts as published on this site: one anonymized industrial production deployment, live since January 2026, 10 critical assets, 5 active production users, daily cadence, approval-gated workflows. The four measured figures and the workflow each was measured on are stated on the value model page. The business-hours clock behind the quote-to-order figure is stated in the maintenance benchmark provenance ledger.
Where to go next
- Water Utility Software Integrations lists what AIMMS connects to and which direction data moves at each connection.
- Autonomous AI Without Plant Control sets out how to test the read and write boundary in any product, including this one.
- How to Measure a Maintenance AI Pilot covers what to instrument before a pilot, so a vendor figure is not the only number you have.
- The Maintenance Benchmark Provenance Ledger grades widely quoted maintenance benchmarks, and grades EQUA’s own four figures by the same rule.