Vibration and Condition Monitoring Software: Custom Build Versus Off the Shelf
Buy. If you are single vendor end to end with under about 200 monitored machines, Emerson or SKF will serve you well and a build reproduces features you already have. Buy first if you have no measurements at all, because software does not create data.
On this page
Buy. If you are single vendor end to end with under about 200 monitored machines, Emerson or SKF will serve you well and a build reproduces features you already have. Buy first if you have no measurements at all, because software does not create data. Build when analyser hardware from several vendors, a protection rack and an oil laboratory each hold part of the same machine's story.
What the off-the-shelf products actually do well
Give the incumbents their due, because they are strong inside their own boundaries. Emerson AMS Machine Works is a capable route and analysis system built around CSI analysers. SKF @ptitude Observer is a mature database and it assumes SKF collection hardware. Baker Hughes Bently Nevada System 1 is the right answer for continuously monitored turbomachinery on 3500 series protection racks, and it is not attempting to be a route database for nine hundred pumps. Augury has a genuinely capable model running on Augury sensors across the asset types it has been trained on.
The measurement layer is not the thing to rebuild either. Accelerometers, analysers, wireless sensors, the protection rack: that is instrumentation with certification behind it, and in hazardous areas it is an electrical engineering project rather than a software one. Oil analysis laboratories do work no build replaces.
Buy, and stop reading here, if you run one vendor's hardware end to end with under roughly 200 monitored machines. A custom platform would be an expensive route to features already sitting in your licence. Buy first, too, if your real problem is that you have no data: install sensors and a vendor platform, run it for a year, then decide. We have talked plants out of a build on exactly that basis.
Where they stop
The failure is not a weak product. It is that a real plant runs two or three of them plus a laboratory, and nobody sells the joining layer because every vendor's commercial interest is that you standardise on them.
Machine identity is the first break. Your route database calls it MILL-DRIVE-2-NDE. The protection rack calls it point 4B on rack 12. The computerised maintenance management system (CMMS) calls it a functional location with a code your maintenance team invented in 2011. Spectra export in incompatible formats, sometimes as a proprietary binary, sometimes as a flat file whose header changes between firmware versions. Until one asset and machine train model exists that every source maps into, with bearings identified by part number so defect frequencies are computed rather than guessed, every analytics project produces a dashboard nobody trusts.
Alarm limits are the second. ISO 20816 gives evaluation zones for broadband vibration by machine class and support type, and it is a reasonable starting point rather than an answer for your machine. A pump on a stiff concrete plinth and the same pump on a skid behave differently. A variable speed drive breaks fixed frequency bands entirely, because the bearing defect tone moves with shaft speed and a static band either misses it or screams continuously. So analysts turn limits up until the noise stops, and the system becomes decorative.
The third break is the handoff. An analyst writes: high 1x with harmonics, suspect misalignment, recommend laser alignment check at next opportunity. That reaches the planner as a notification with the text pasted in, the planner raises a job called check vibration, a fitter finds nothing obviously wrong and closes it, and the coupling fails three months later. Nobody was negligent. The diagnosis lost all of its specificity in transit, and no packaged system carries fault type, confidence, affected component, scope and tooling into a maintenance notification with the right catalogue codes.
The arithmetic per monitored machine
Price this per monitored machine per year, because that is the unit your programme is budgeted in. Substitute your own contract figures.
Suppose your condition monitoring platform costs $90 per monitored machine per year across 900 machines, so about $81,000 annually, before the analysers themselves and before the laboratory subscription. A first release covering the unified machine train model, ingestion from two or three sources, an alarm engine you control, a triage queue and notification creation straight into the CMMS runs $60,000 to $130,000 in our delivery experience and ships in 12 to 16 weeks. Take $95,000, amortise across five years, add year two support at 15 to 20 percent, and you land near $36,000 a year.
On licence alone that puts the crossover near 400 monitored machines. Treat that number carefully. It is only real if the build lets you retire one of the vendor databases, and if the retired system is a protection rack, it does not, because the rack is a safety function and it stays. In practice the crossover is around 500 machines and it only bites when you carry two or more vendor platforms at once.
The larger number is the analyst. If a person spends most of a week collecting and a fraction of it analysing, you are funding data gathering rather than diagnosis, and route intervals driven by consequence rather than by a fixed calendar are what free that time without cutting coverage where it counts.
What a custom build actually costs
The first release above runs $60,000 to $130,000 across 12 to 16 weeks, and it is a working system for one plant rather than a proof of concept. Adding raw waveform and spectrum storage with analysis in the browser, order tracking for variable speed assets, oil analysis and thermography ingestion, criticality driven route scheduling, mobile collection support, closure reporting and machine learning triage runs $150,000 to $400,000 over 6 to 12 months.
Data migration runs 10 to 25 percent of the build, and this category sits at the top of that band whenever you want history. Reading three years of trends and spectra out of proprietary databases and mapping them onto a new point identity is the migration, and it needs deciding up front rather than in month four. Year two costs 15 to 20 percent of the build annually, driven by firmware changes to export formats and by CMMS upgrades.
What pushes cost up: raw waveform volume, since storing time waveforms for nine hundred points monthly is a genuine storage and query design problem. Proprietary export formats, because reverse engineering one vendor's binary is contained work and doing it for three is not. Historian integration, particularly where speed and load context comes from a control system through an interoperability layer with its own security review. And hazardous area constraints on new wireless sensors. What keeps it down: one plant, the top hundred machines by consequence, and two data sources.
The four situations where building wins
- Regulatory fit. Where condition monitoring evidence feeds a mechanical integrity programme, an insurer's inspection regime or a safety case, the record has to be producible and traceable rather than living in a vendor database an analyst exports from. Personnel certification under ISO 18436 governs who may make a call, and recording who made it, against what evidence, is part of the same obligation.
- Scale economics. Past roughly 500 monitored machines carried across two or more vendor platforms, the licence lines plus the duplicated analyst effort outrun the build. Below that, and especially with one vendor, the licence wins.
- A workflow that is your competitive advantage. For a plant this reads as availability rather than market position. If your programme's hit rate is the reason a capital case for spares or a shutdown scope gets approved, being able to state that hit rate is worth owning. Most programmes cannot state it at all, which is why they get cut in a bad year.
- Integration sprawl across three or more systems. Route database, protection system, oil laboratory, thermography folder, historian and CMMS. Six sources telling the truth about the same gearbox, and no human with a reason to open all of them on the same day for the same asset. That is the failure that produces a Sunday seizure with the evidence already in the building.
How to decide in a week
Run the hindsight assembly. Pick a machine that failed unexpectedly in the last two years. Give an engineer one day and ask for every piece of evidence that existed in the eight weeks before it went down: route spectra and trends, protection system data if it was wired, oil sample results, thermography images, and any work order raised against it.
Then count. How many sources had to be opened, how long it took, and how many of those signals were individually inside their alarm limits while collectively pointing the same way. In most plants the answer is that nothing was hidden and nobody had a reason to look at all four on the same day, which is precisely the case for a joining layer.
Run the second test forward. Take last month's alarms and ask your analysts how many they acted on, how many they dismissed, and how many they would have dismissed regardless of the reading. If limits have been raised until the noise stopped, your alarm system is decorative and the fix is order normalised bands and statistical limits from each machine's own baseline, not more sensors.
Third, take ten diagnoses from last quarter and trace each to its work order and its outcome. If you cannot say whether the call was right, you cannot state a hit rate, and you cannot defend the programme's budget.
Then move to a paid discovery phase rather than a proposal. Two to three weeks with your reliability engineer and a planner, ending in a signed product requirements document covering the machine train model, the point identity mapping, waveform retention policy with numbers attached, CMMS notification codes and acceptance criteria. That specification is yours whichever firm builds. Digital Heroes writes one before code exists, gives you a named team to meet first, and is checkable on Clutch, Trustpilot, Fiverr Vetted Pro and D-U-N-S. We are the wrong firm for a plant with 150 machines on one vendor's analysers.
Book a 30-minute call with Digital Heroes and get a written plan and a fixed quote within 48 hours.
The evidence behind this guide
Independent findings on why this investment pays off. Every link goes to the primary source.
- McKinsey found that tech debt can amount to 20-40% of the value of a company's entire technology estate before depreciation, and CIOs report that 10-20% of the budget for new products is diverted to resolving tech-debt issues. Source: McKinsey & Company (2020) →
- Standish's 2015 CHAOS research found roughly a third of software projects (about 36% by the Modern definition) fully succeed on time, on budget, and on scope, with top success drivers including executive support, user involvement, and clear requirements/business objectives. Source: Standish Group (CHAOS Report) (2015) →
- 48% of private companies cite integration with legacy systems or technical debt as a top obstacle to realizing the full value of their digital and AI investments (behind data quality/availability at 72% and gaps in AI fluency or technology talent/leadership at 53%). Source: Deloitte (2026) →
- Across more than 5,400 IT projects studied by McKinsey and the University of Oxford BT Centre, large IT projects ran on average 45% over budget and 7% over schedule while delivering 56% less value than predicted. Source: McKinsey & Company / University of Oxford (BT Centre for Major Programme Management) (2012) →
Frequently asked questions
How much does custom condition monitoring software cost?
A first release with a unified machine train model, ingestion from two or three sources, an alarm engine you control, a triage queue and maintenance notification creation runs $60,000 to $130,000 in 12 to 16 weeks in our delivery experience. Adding waveform storage with browser analysis, order tracking, oil and thermography ingestion, route scheduling and closure reporting runs $150,000 to $400,000 over 6 to 12 months.
Do we have to replace our Bently Nevada protection system?
No, and you should not. A 3500 series protection rack performs a machinery protection function with its own safety and certification standing, and it stays exactly where it is. What a build adds is a layer that reads from it alongside your route database, oil results and thermography, so a turbomachine and a pump can be reasoned about in the same place. Replacing protection hardware is an instrumentation project, not a software one.
How do you set alarm limits that analysts will actually trust?
With bands defined as multiples of running speed rather than fixed frequency, so a variable speed asset does not defeat them, and with statistical limits derived from each machine's own baseline period rather than a class table. Add rate of change alarms, because a defect amplitude that doubles in two weeks matters more than one slightly raised for three years. ISO 20816 zones are a starting point, not an answer for your specific machine.
Who owns the code and the waveform data if we commission a build?
You should own the repository, the infrastructure accounts and the data, in writing before kickoff. At Digital Heroes the client owns all of it from the first commit and is free to hire anyone else to continue. This point is sharper here than elsewhere: the entire reason to build rather than buy in this category is escaping hardware vendor lock, so accepting a new lock from your developer would defeat the exercise entirely.
What happens if a vendor firmware update changes the export format?
The ingestion layer should fail loudly and quarantine the source rather than silently importing wrong data, and the parser should be versioned so you can support two formats during a rollout. Ask any developer which vendor exports they have parsed, by product and version, and what they did when a header changed. That question separates people who have done this from people who have read a specification.
Can we start with one plant and expand later?
Yes, and it is the cheapest sequence. Start with one plant, the top hundred machines by consequence, and two data sources. The machine train model and point identity mapping built there carry to the next site, so the second plant costs less than the first. Trying to model every asset at every site before anything ships is the reliable way to turn a 16 week release into a year of workshops.
What is the difference between condition monitoring and predictive maintenance?
Condition monitoring is the measurement and diagnosis activity: collecting vibration, oil, thermal and process data, then judging machine health from it. Predictive maintenance is the maintenance strategy that acts on those judgements by scheduling work before failure. Software vendors use the terms interchangeably, which matters because a tool that produces alerts without a route into a work order with the correct scope has done only the first half.
Is machine learning worth building into a condition monitoring system?
Not in year one. Anomaly detection without labels produces many alerts and no diagnosis, and the useful models are trained on your own confirmed outcomes. Start collecting labels the day the system goes live by forcing every diagnosis to a closure record with the actual finding, and the model becomes worth funding in year two. Any vendor selling a model that needs no labelled outcomes from your plant is selling optimism.
How do we prove the programme is working when budgets get cut?
By stating a hit rate you can defend, which requires closing the loop. Every diagnosis needs an outcome field completed when the work is done, with the actual finding, photographs and the failure mode confirmed or corrected. Most programmes cannot answer how many calls were correct or how many failures were missed, and a programme that cannot answer those two questions is judged on whether it caught the last big one.
How much waveform history should we keep?
Decide it before the build rather than after, because retention drives the storage and query design. Storing time waveforms for nine hundred points monthly across three years is a real engineering decision, not a rounding error. A common compromise keeps full waveforms for a rolling window, downsampled spectra and trends indefinitely, and full history on the machines whose consequence justifies it. Put the numbers in the specification.
When does a company outgrow Airtable?
The usual breaking points are record limits, permissions, and automation complexity. Airtable's Team plan caps each base at 50,000 records and Business at 125,000, so operations logging thousands of rows a month hit the ceiling within a year or two. The other trigger Digital Heroes sees constantly is permissions: restricting who can view specific fields or records is clumsy below Airtable's Enterprise tier, which becomes a genuine problem once salaries, pricing, or client contracts live in the base.
How many developers does it take to build an internal tool?
Two to four people covers nearly every internal tool: one or two developers, a part-time designer, and a project manager who doubles as your single point of contact. Internal tools rarely need consumer-product polish, so a full-time dedicated designer is usually wasted budget. On Digital Heroes projects, a two-person core team handles the typical 4 to 8 week build, with a specialist pulled in briefly for a tricky integration or a security review.
How do I calculate the ROI of a custom internal tool?
Count hours first: multiply the weekly hours staff spend on the manual process by their loaded hourly cost, then add the cost of errors such as mispriced quotes or missed renewals. A tool saving a 10-person team 5 hours each per week recovers about 2,500 hours a year, which repays a $20,000 to $30,000 build well inside a year at typical wages. Most internal tools Digital Heroes delivers reach payback in 6 to 18 months, with quoting and billing tools at the fast end because they plug revenue leaks, not just time.
How many SaaS seats do we need before building custom becomes cheaper?
The crossover usually shows up between 20 and 50 seats on premium tiers. Salesforce Enterprise lists at $165 per user per month, so 40 users cost about $79,000 a year in subscriptions, which is real money against a custom system you would own outright. Run the comparison over three years: if subscription spend beats the build cost plus 15-20% annual maintenance, custom wins on price before you even count workflow fit.
Should we build our internal tool in Retool instead of hiring developers?
Retool is the right choice if someone on your team is comfortable with SQL and JavaScript and the audience is a handful of technical users, because a basic CRUD dashboard comes together in days. Hire developers when non-technical staff will use the tool daily, when the logic goes beyond forms sitting on a database, or when per-seat pricing stings, since Retool's Business tier lists at $50 per standard user per month. A pattern Digital Heroes sees often: companies arrive after a year on Retool with a tool nobody can maintain because the one person who built it has left.
What questions should I ask a development agency on the first call?
Ask who exactly will build it, what happens when scope changes mid-project, what their maintenance terms are after launch, and what they will need from you every week. Then ask them to describe a project that went wrong and what they changed afterward; teams that have shipped at real volume have war stories, and teams claiming a perfect record are hiding something. The scope-change answer matters most: a disciplined shop describes a written change-order process, not a vague promise to be flexible.
Can I build my product on a no-code tool like Bubble instead of hiring developers?
For testing whether anyone wants the product, yes, and Bubble's paid plans start at $29 a month, which is the cheapest validation you will ever buy. The ceiling arrives with complex data relationships, heavy integrations, performance at a few thousand users, and the fact that you cannot export a Bubble app to servers you control. A path many Digital Heroes clients take: prove demand on no-code, then rebuild custom once revenue justifies it, treating the no-code version as a paid prototype rather than a foundation.
Should I hire a freelancer or an agency for my software project?
A skilled freelancer is the right call for a single-discipline scope under roughly $15,000, like a website, a plugin, or one integration. Above that, projects need design, backend, testing, and project management at once, and a solo builder becomes the single point of failure: if they get sick or take a bigger client, your project simply stops. Agencies bill 20-40% more per hour but carry continuity, code review, and someone to escalate to, which is what you are actually buying.
How do we migrate years of spreadsheet or Airtable data into a new internal tool?
Migration is a standard part of the build, not a separate project: the agency writes import scripts that clean, deduplicate, and map your existing rows into the new database. On typical spreadsheet and Airtable histories, Digital Heroes budgets 3 to 10 extra days, most of it spent resolving inconsistencies like the same customer spelled four different ways. The safe sequence is a trial migration first, a review of flagged conflicts with your team, then final cutover over a weekend so nobody loses a working day.
Who can build a custom internal tools system?
Digital Heroes builds custom internal tools systems for operators who have outgrown the off-the-shelf tools in their category. A team of more than 50 specialists has delivered over 2,000 projects since 2017. Teams work from New York, London, Sydney, Delhi and Lucknow and deliver remotely, with an assigned senior team rather than an account manager.
Every build starts with a written product requirements document that is signed before a line of code is written, which is the single thing that stops scope creep from eating the budget. Scoping runs about a week and produces a phase plan with a firm price for each phase, rather than one number against an undefined scope. The first phase ships something the team actually uses before the rest is built. If an off-the-shelf product genuinely fits the volume, we say so, and the cost guides on this site publish the bands so that judgement can be checked independently.
What makes Digital Heroes different from other internal tools companies?
Four things that competitors in this bracket cannot simply copy. Digital Heroes runs a YouTube channel with more than 2.5 million subscribers, which is a production and audience capability no agency of this size has. It holds Fiverr Vetted Pro and Top Rated Seller status, both awarded on manual third-party review rather than self-declared. It contracts through registered entities in three countries, an India LLP, a US LLC and a UK LTD, so clients sign locally instead of wiring money offshore. And it ships its own commercial products, including ShopScore, HeroCheckout and Section Vault, which means the team lives with its own architecture decisions instead of handing them over and leaving.
Two more that show up in the work. Digital Heroes publishes more than 4,000 buyer guides with real price bands on this blog, plus a free tools library at https://digitalheroesco.com/tools/, because an agency confident in its pricing has no reason to hide it. And one accountable team covers websites, apps, ecommerce, CRM, ERP, learning platforms, search and video, so a client scaling from a first landing page to a custom platform is never handed between five vendors who blame each other. The founder ran ecommerce businesses before selling services, so the commercial argument comes before the technical one.
How can I check Digital Heroes is legitimate before getting in touch?
Verify it independently rather than taking the site's word for it. The YouTube channel is at https://youtube.com/@DigitalMarketingHeroes, the Fiverr profile at https://www.fiverr.com/shreyanshsin261, and the Upwork profile at https://www.upwork.com/freelancers/shreyanshsingh. Client reviews sit on Clutch at https://clutch.co/profile/digital-heroes-0 and Trustpilot at https://www.trustpilot.com/review/digitalheroes.co.in, and the company page is at https://www.linkedin.com/company/digital-heroes-1/.
Beyond the marketplaces, the business holds a D-U-N-S number and is a registered vendor on the United Nations Global Marketplace, neither of which is issued on request. Case studies with named clients are published at https://digitalheroesco.com/case-studies/. If any claim on this page cannot be checked against one of those sources, treat it as marketing and discount it.
Related guides
More from Ahaan.
Senior Android Engineer · Delhi · 1 post on the site
Published · Last updated .
