idoc PredictIT
idoc PredictIT is predictive maintenance software for instrumented industrial equipment. It builds a statistical model of normal operation from sensor data alone, then raises a surprise index alarm naming the measurement points that deviate. Cloud or local network deployment.
What is idoc PredictIT?
idoc PredictIT is predictive maintenance software for industrial equipment, published by idoc A/S, a Danish engineering company based in Aalborg and registered under CVR number 32838650 since 1 April 2010. The publisher describes itself as an engineering company and a knowledge house within the engineering industry, and presents PredictIT as its predictive maintenance offering. The product was, in the publisher's words, developed in collaboration with HUGIN, a Bayesian network software company.
The technical principle is stated plainly. A statistical model automatically estimates the correlations between an item of equipment's measurement points while that equipment is running normally. Because the model describes healthy behaviour rather than catalogued faults, no database of labelled breakdowns is needed, which matters where failure history is thin or absent. Once in production, the model receives measurement data continuously and computes a surprise index comparing incoming values with expected behaviour. A surprising value raises an alarm to an operator and, importantly, specifies which measurement points are behaving abnormally, so the source of the deviation is identified rather than merely signalled. The vendor states that the tool is based on advanced AI/ML technology combining knowledge and experience with data.
idoc claims transparency for the model: its credibility can be assessed by users or experts, and it can be adjusted with operators' knowledge and experience. The same model can be reused in other contexts, for instance troubleshooting a system that has already failed. The tool is also framed around spare parts, with the stated vision of getting them to the right place at the right time before a critical breakdown, and around detecting patterns in unplanned maintenance events.
Implementation follows four named steps, Data collection, Model Construction, Integration and idoc PredictIT is live, and the solution can be implemented in the cloud or locally on the company's own internal network.
What the site leaves out matters as much. It runs to five pages in total, with no features page, no technical documentation, no public demo, no product API, no mobile app, no named customer and no published price. idoc PredictIT is a contracted engineering solution sold through a sales conversation, not a product you can evaluate on your own.
What it does
- Estimate a statistical model of the correlations between measurement points automatically, from data recorded during normal operation.
- Compute a surprise index on incoming data and detect values that depart from expected behaviour.
- Raise an alarm to an operator and name the measurement points that are behaving abnormally.
- Connect an item of equipment's measurement points to idoc PredictIT for continuous monitoring.
- Validate and adjust the model with software tools before it goes live.
- Refine the model with the knowledge and experience of the operators who run the equipment.
- Plan the supply of spare parts ahead of a critical breakdown.
When to use idoc PredictIT / When not to
A quick filter to help you decide if idoc PredictIT is the right fit.
When to use idoc PredictIT
- Operators of instrumented industrial equipment whose measurement points already produce usable data.
- Maintenance, reliability and spare-parts teams moving from corrective or calendar-based servicing to condition-based monitoring, so that parts reach the right place before a critical breakdown.
- Plants with a thin or non-existent labelled failure history, since the model is estimated on normal operation rather than on catalogued breakdowns.
- Engineering teams that want a transparent model they can audit and enrich with operator knowledge, rather than a black box.
- Organisations that need process data to stay in-house, since the solution can be installed locally on the company's own internal network.
When not to use idoc PredictIT
- Buyers who need public pricing or self-service onboarding: there is no price list, no sign-up flow, and neither a free trial nor a free plan is advertised, so the only way in is a sales conversation.
- Developers expecting a product API or a mobile app: none is published, API route probes returned genuine 404s, and App Store and Google Play lookups came back empty against working controls.
- Operators of equipment without usable measurement points, since the very first step requires a sensor dataset recorded during normal operation.
- Evaluators who want to assess the product on their own: the site has five pages in total, with no technical documentation, no features page and no public demo.
- Teams that need a documented interface or contract language other than English: no interface language is published, and the terms and conditions are interpreted and construed exclusively in English.
How to use idoc PredictIT
A typical end-to-end flow, from setup to results.
- Get in touch with idoc through the Contact us call to action carried on every page, or the Contact Sales email address on the contact page; there is no public sign-up path.
- Check that the equipment you want to monitor already has measurement points producing usable data.
- Step 1, Data collection: make a relevant dataset available, recorded from those measurement points while the equipment runs normally.
- Step 2, Model Construction: the dataset is used to estimate a model of the correlations between measurement points automatically.
- Validate the model and adjust it where necessary using software tools.
- Step 3, Integration: connect the equipment's measurement points to idoc PredictIT.
- Settle the deployment architecture: cloud, or a local installation on the company's own internal network.
- Step 4, idoc PredictIT is live: the model receives data continuously and computes the surprise index.
- Act on alarms: an operator is alerted and the deviating measurement points are identified for investigation.
- Feed operators' knowledge and experience back into the model, and reuse it in other contexts such as troubleshooting a system that has already failed.
Pros & Cons
Pros
- No labelled failure history required: the model is estimated on normal operation, not on a catalogue of past breakdowns.
- Works from sensor data that is already available, described by the vendor as a model built from sensor data only.
- The model is presented as transparent and its credibility as assessable by users or experts, rather than a black box.
- Alarms name the offending measurement points instead of emitting a binary signal, which shortens investigation.
- Operators' field knowledge can be injected into the model, and the model can be reused in other contexts.
- Local installation option: industrial data can stay on the customer's own internal network.
- Backed by an established publisher, idoc A/S, registered since 2010 and listing 51-200 employees on LinkedIn, with HUGIN as a Bayesian network partner.
Cons
- No public pricing whatsoever: no price list, no order of magnitude, not even a pricing page, on a site that totals five pages.
- No free trial and no free plan is announced, and no self-service entry point exists.
- No public technical documentation, no features page and no demo, so the product cannot be evaluated remotely.
- No product API and no mobile app: API route probes returned genuine 404s, and iTunes and Google Play lookups were negative against working controls.
- No security certification is published, neither ISO 27001 nor SOC 2 nor any other.
- No data processing agreement, no sub-processor list and no named hosting country; the privacy policy covers the website only, not the industrial data entrusted to the tool.
- No named customer, no performance figure and no quantified use case; unedited template placeholders are still present in the footer links.
Pricing & Plans
No price is published. idoc PredictIT has no pricing page: three exhaustive enumerations of the site, through its page sitemap, its llms.txt file and the WordPress REST API, return the same five pages, none of them commercial. A search of the collected pages for price, pricing, currency, subscription, free trial or demo terms returns no amount at all. No free trial, no free plan, no credit pack and no commitment is described anywhere. The publisher's own site lists Competetive Prices (sic) among its selling points without attaching a single figure to it. Pricing is therefore established case by case: a quote is obtained by contacting idoc through the Contact us call to action carried on every page, or through the Contact Sales email address published on the contact page. This is why the starting price, currency and billing unit fields are left empty on this record.
Data, GDPR & hosting
A consolidated view of how idoc PredictIT handles your data.
GDPR overview
Compliance is claimed in writing: idoc states that its processing of personal data is consistent with the applicable legislation, such as the General Data Protection Regulation. The policy names its legal bases, Article 6(1)(b) and Article 6(1)(f), and lists data subject rights with their articles: access (15), rectification (16), erasure (17), restriction and objection (21). Rights are exercised at Info@idoc.dk, and the supervisory authority named is the Danish Data Protection Agency (datatilsynet.dk). The gaps are just as visible: no data protection officer, no published data processing agreement, no sub-processor list and no described transfer mechanism outside the EEA. As a Danish company, idoc needs no Article 27 representative. One drafting error: the right to restriction is attributed to Article 16 instead of Article 18. The policy covers the website only, never industrial sensor data.
Who owns the data?
The published documents cover the website, not the plant. Clause 4 of the terms and conditions, last updated on 13 January 2025, reserves the intellectual property in the site and its resources to idoc or its licensors. Clause 7 goes further: any idea submitted without a prior written agreement grants idoc a worldwide, irrevocable, non-exclusive, royalty-free licence to use, reproduce, store, adapt, publish, translate and distribute that content. The privacy policy names idoc as responsible for processing personal data concerning its business partners and customers. Nothing published states who owns the sensor and process data a customer entrusts to idoc PredictIT, or what idoc may do with it; that has to be settled in the contract.
Reuse rights
The terms grant the end user no reuse right: clause 4 reserves the site and its resources to idoc and its licensors, while clause 7 runs the other way and licenses to idoc any idea a visitor submits without a prior agreement. On personal data, the privacy policy relies on Article 6(1)(b) of the GDPR (performance of a contract) and Article 6(1)(f) (legitimate interest). The data described is business contact data: company name, CVR number, contact person, telephone, email, website and payment information. Cookies collect a unique identifier, technical information about the device, IP address, geographic location and pages clicked, for technical, statistical, personalisation and marketing purposes, with consent gathered through a banner. idoc states that it entrusts website personal data to data processors, but publishes no list of them. Whether customer or sensor data is used to train models is never addressed, in either direction.
Data retention & training
Hosting summary
The only statement about where the product runs is that the solution can be implemented in the cloud or locally on the company's own internal network. No country, no region, no cloud provider and no data centre is named. No sub-processor is identified either: the privacy policy says idoc entrusts website personal data to data processors, without listing them, and describes no mechanism for transfers outside the EEA. The publisher is established in Denmark, so its own processing falls under Danish and EU law, but that follows from where the company sits, not from any hosting commitment. One measured fact concerns the marketing site alone: idocpredictit.dk resolves to 94.231.106.67, announced by AS48854 team.blue Denmark A/S in Aarhus, with no anycast flag, so the website is served from Denmark. That says nothing about where the platform would process sensor data. Organisations with a data residency requirement should note that the local installation is the only published way to keep industrial data inside their own network, and should have hosting, sub-processors and transfers written into the contract.
Things to keep in mind
Risks and trade-offs to weigh before adopting idoc PredictIT.
- Template residue in the footer of every page: the links point to mailto:contact@mysite.com and tel:123-456-7890 while the displayed text reads joe@idoc.as and +45 35 12 39 39. The displayed details are the real ones, the phone number matching the domain registrant record exactly; the links are unedited Elementor placeholders, and a sign the site has not been fully proofread.
- Two divergent telephone numbers: 20 60 50 82 on the contact page, +45 35 12 39 39 in the footer. Only the second appears in the Danish company register and the domain whois record, so confirm which line to call before relying on either.
- Three domains for one publisher: idocpredictit.dk for the product, idoc.as for the publisher's own site and its sales address, idoc.dk for the GDPR contact. No postal address is published on the product domain itself.
- HUGIN is a development partner in Bayesian networks, not an integration: developed in collaboration with HUGIN says nothing about connectors a customer can switch on. No third-party connector is named anywhere, which is why the integrations field is left empty.
- The privacy policy covers the website only. It says nothing about the sensor data entrusted to the tool: no ownership clause, no retention period, no hosting location, no position on model training. It also contains a legal typo, attributing the right to restriction of processing to Article 16 instead of Article 18 of the GDPR.
- No certification is claimed anywhere, neither ISO 27001 nor SOC 2 nor any other, so none should be assumed. Likewise, the WordPress REST endpoint at /wp-json that answers with HTTP 200 is CMS plumbing, not a product API: probes of /api, /api/v1 and /api/v1/health returned genuine 404s and the site never mentions an API.
- Part of the company identity comes from off-site registers rather than the website: the 1 April 2010 start date is from the Danish CVR register, and the legal entity has changed name twice under the same CVR number (IDOC ENGINEERING ApS 2010-2012, IDOC ApS 2012-2015, IDOC A/S since 6 October 2015). Its registered purpose is generic, handel og investering under NACE code 829900, although company name, address, telephone and email all match the site.
Setup & Integrations
Technical difficulty
High for a self-service buyer, moderate for an already instrumented plant. This is an engineering project run with the publisher, not a sign-up: the equipment must carry usable measurement points, a dataset has to be recorded during normal operation, the resulting model is validated and adjusted with software tools, and the measurement points are then connected to idoc PredictIT. One architecture decision comes up front, cloud or local installation on the internal network. No public technical documentation, connector list or protocol is published, so effort and lead time can only be established with idoc.
Behind idoc PredictIT
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
How does idoc PredictIT detect a problem?
Do I need a history of past failures to use it?
What data do I have to provide?
What happens when an anomaly is detected?
Can the solution run inside our own network?
Is the model a black box?
Who publishes idoc PredictIT?
How much does it cost?
Is there an API or a mobile app?
How do I exercise my GDPR rights?
Should you pick idoc PredictIT?
idoc PredictIT is a niche tool sold under contract by an engineering company, not a self-service SaaS, and it is best read as the software arm of a predictive maintenance practice. Its technical differentiator is coherent and unusual enough to justify a conversation: the model is estimated on normal operation instead of a catalogue of past failures, alarms name the deviating measurement points instead of emitting a binary signal, the model is claimed to be transparent and assessable by users or experts, and operator know-how can be fed back into it. For a plant that is already instrumented but short on labelled failure history, that combination addresses a real obstacle.
The commercial side is entirely opaque. No price, no order of magnitude, no free trial, no named customer and no published metric: five pages that document the principle rather than the delivered product. Nothing tells you what the platform looks like day to day, what it connects to, or where it runs.
The publisher, by contrast, is solid and verifiable: idoc A/S, Aalborg, CVR 32838650, continuously registered since 2010, with a stated development partnership with HUGIN on Bayesian networks.
If the tool fits your situation, bring a written list to the sales call: the price and how it is calculated, what the integration step actually connects to and over which protocols, who owns the sensor data and how long it is kept, where the platform is hosted if you take the cloud option, whether a data processing agreement is available, and exactly what the local installation covers. The published privacy policy answers none of these for industrial data. It covers the website only.
- Choosing a selection results in a full page refresh.
- Opens in a new window.