Industry agnostic
Applicable across all heavy asset industries.
Data Standardisation and AI Readiness Framework
Standardised asset data, trusted decisions, AI-ready foundations.
A maintenance or investment decision is only as good as the records it is read from. Standardising asset data is the work that makes those records dependable, at every site and in the same terms.
This framework is how Shivaan Asset Management does that work. Five phases and nine steps take an asset base from records kept a different way at every site to standardised data that reporting, optimisation and AI can stand on. It is written to be learned and applied by anyone, in any heavy asset industry.
Structured data is the prerequisite for every capability that follows, not the afterthought once the tools are bought. Each step below sets out what it is, why it matters and the value it returns.
In brief
- The framework standardises asset data through nine steps, grouped into five phases: Discover, Structure, Standardise, Govern and Activate.
- It applies in any heavy asset industry, and each step is engineered from reference standards and methods like ISO 14224, RCM and FMEA rather than local habit.
- Discover maps and classifies the asset base. Structure decomposes each asset and standardises the language of its maintainable items.
- Standardise builds a reusable RCM knowledge library and generates CMMS-ready failure codes. Govern keeps the data owned and improving, and Activate switches on dashboards, analytics and AI on data you can trust.
- Structured data is the prerequisite for reliable reporting, optimisation and AI, not an afterthought.
Five principles behind the method
These hold true at every step, in any heavy asset industry.
1 2 Proven approach
Engineered for real-world complexity.
3 Reusable and scalable
Create once, use many, grow with confidence.
4 Governed and secure
Trusted data with clear ownership and control.
5 Deliver value
Outcomes the business can see and measure.
The five phases and nine steps
The phases run in sequence. Each one makes the next possible, so nothing downstream is built on data that has not been agreed. Every step sets out what it is, why it matters and the value it returns.
01
Discover
Know what you have, and classify it by design, before anything else.
What it leaves in place. The asset scope is mapped and organised into a consistent, three-level classification that every later step reuses.
01
Understand Asset Portfolio
- Asset register
- Operating context
- Functional boundary
Map every physical asset: what exists, where it operates, who owns it and the function it performs, with its operating context and functional boundaries.
- Why it matters.
- Every downstream step depends on this baseline being accurate and complete, so nothing is assumed and no inherited register is taken at face value.
- The value.
- One dependable view of the asset base that the rest of the framework is built on.
02
Classify Asset Types
- Asset class
- Asset class type
- Variation
- ISO 14224
Organise every asset into three levels: asset class, asset class type, and the variations where design differences change how an asset fails or is maintained.
- Why it matters.
- Classifying by design, not just by function, is what makes failure knowledge reusable, because design and operating context drive failure behaviour.
- The value.
- A consistent, three-level asset language that every later step reuses.
02
Structure
Decompose each asset to the component, then standardise the language of its maintainable items.
What it leaves in place. A complete engineering breakdown and a shared, controlled vocabulary of maintainable items that reads the same at every site.
03
Build Equipment Hierarchies
- System to component
- Maintainable item
Decompose each asset class type from the system level down to the component, using engineering logic: system, sub-system, assembly, sub-assembly, component.
- Why it matters.
- How far you decompose is set by design complexity and maintenance significance, not by habit or by what another industry did.
- The value.
- An engineering breakdown that locates every maintainable item in the asset.
04
Standardise Maintainable Items
- Maintainable items
- Ontology
- Nomenclature
- CMMS-compatible
From the equipment hierarchy, extract and standardise every maintainable item and component under controlled nomenclature, so the same thing is named the same way everywhere.
- Why it matters.
- Standardised names are what make asset data reusable, searchable and CMMS-compatible across sites, asset classes and enterprise systems.
- The value.
- A shared vocabulary of maintainable items that unlocks interoperability, today and with every future system.
03
Standardise
Turn failure knowledge into a reusable library, then into load-ready CMMS failure codes.
What it leaves in place. A templated RCM knowledge library and CMMS-ready failure codes, structured to deploy at scale across any site.
05
Create Reusable Knowledge Library
- RCM
- FMEA
- Failure mechanisms
- Failure causes
Develop a reusable engineering library of failure modes, mechanisms, causes, effects and maintenance task logic, captured at component level in standardised terms.
- Why it matters.
- Once built, the library is templated and reused across asset classes and sites, so failure knowledge becomes the organisation's most durable asset, outlasting any individual engineer.
- The value.
- A templated failure-knowledge library you build once and deploy many times.
06
Generate CMMS-ready Failure Codes
- PCR
- SAP PM
- Maximo
- Load-ready
Convert the library into structured failure code sets, Problem, Cause, Remedy (PCR), formatted for the target CMMS such as SAP PM or Maximo, with every record traceable to a failure mode in the library.
- Why it matters.
- Codes built this way are designed to load first time: reviewed and approved on first pass, minimising rework, gaps and free-text drift.
- The value.
- Load-ready failure codes your CMMS accepts and your analytics can trust.
04
Govern
Close the loop so the data improves itself, and give it clear, lasting ownership.
What it leaves in place. A self-sustaining improvement system, owned and stewarded independently of any one person.
07
Improve Data Feedback Loop
- Governance
- Feedback loop
- Continuous improvement
Establish the governance cycle that feeds work order history, condition monitoring and incident records back into the library.
- Why it matters.
- Data quality compounds with every maintenance cycle: the library gets smarter with every failure, turning operational experience into reusable engineering knowledge.
- The value.
- A library whose accuracy improves the longer you run it.
08
Steward Ownership and Governance
- Ownership
- Review cadence
- Sustainability
Assign clear ownership, set a review cadence, and embed the framework into standard operating practice across operations, maintenance and engineering.
- Why it matters.
- The process has to be sustainable independently of any individual, built to outlast its creators and improve as the organisation matures.
- The value.
- A durable, owned, cross-functional way of working, not a one-off project.
05
Activate
Activate dashboards, analytics and AI on data you can finally trust.
What it leaves in place. Reporting, optimisation and AI decision support running on foundations that were engineered to be AI-ready.
09
Activate Digital and AI Readiness
- Dashboards
- Analytics
- Predictive maintenance
- AI-ready
With standardised data in place, activate the digital layer: dashboards that show meaningful signals, analytics that surface real failure patterns, and AI models that learn from clean, consistent data.
- Why it matters.
- This is the compounding return on every step above. Dashboards, analytics, optimisation and AI decision support are only as credible as the data underneath them.
- The value.
- Predictive maintenance, anomaly detection and optimisation built on foundations that were engineered to be AI-ready.
Outcomes that matter
What standardised, AI-ready asset data delivers back to the business.
Strategic alignment
Asset decisions tied to board-level business priorities.
Improved reliability
Safer, more available and more predictable assets across the portfolio.
Lower maintenance cost
Less waste and more value from every maintenance dollar spent.
Consistent data
Reports and decisions everyone in the business can trust.
Standardisation at scale
One language across every site, asset class and team.
Faster reporting and compliance
Audit-ready evidence and regulatory reports without delay.
People empowered
Teams confident using clean data to make better calls.
Sustainable value
Asset performance and value maintained across the full asset life.
Proof from our projects
Three engagements built on this sequence. Every figure below is the figure already published on that project page.
78%
Failure-code adoption within 8 months, from no prior use of codes
A global mining operation standardised 100+ asset classes and 150+ asset class type variations, then crews began coding failures with the standardised codes and left less of the history in free text.
5.0 / 5.0
Every one of 11 independent client assessment dimensions
A Tier 1 oil and gas operator standardised 5 asset classes and 44 asset class types, ending with 100% standardisation across all deliverables and a Net Promoter Score of 10 out of 10.
2 sites
Coal handling and preparation plant functional locations standardised
An audit across two coal mine plants produced one governed enterprise master data standard covering how functional locations are created, owned and audited, positioning the client's asset management system for ISO 55001 alignment.
Key terms
The vocabulary the framework standardises.
- Maintainable item
- The lowest level in the asset breakdown that is actually maintained, replaced or repaired as a unit.
- Asset class, type and variation
- The three-level classification: the category, its fundamental design architecture, and the design variations that change how an asset fails or is maintained.
- Equipment hierarchy
- The engineering decomposition of an asset from the system level down to the component.
- Ontology
- A controlled, standardised vocabulary for naming maintainable items and components, and the relationships between them, consistently across sites and systems.
- RCM knowledge library
- A reusable library of failure modes, mechanisms, causes, effects and maintenance tasks, captured at component level.
- Failure code (PCR)
- A structured Problem, Cause and Remedy record that turns a failure into data a CMMS can store and analyse.
- CMMS
- The computerised maintenance management system, such as SAP PM or Maximo, that holds asset, work order and failure records.
- ISO 14224
- The international standard for collecting and exchanging reliability and maintenance data, used here as a classification reference.
The foundations, explained
The building blocks the phases rely on, written up in full in our resources.
- Asset Classification: Asset Class, Types and Variations
- Equipment Hierarchy Development: An Anatomy of Asset Knowledge, Intelligence and Standardisation
- Functional Location Hierarchies: A Foundation for Operational and Data Excellence
- Asset Management Policy: The Governance Cornerstone Every Asset-Intensive Organisation Needs
- Asset Criticality Analysis: The Prioritisation Foundation Every Asset Management Decision Depends On
- CMMS Data Quality and Failure Code Integrity: Your Asset Analysis Is Only as Good as Your Records
- Asset Management Maturity: What Each Level Actually Looks Like in Practice
Frequently asked questions
What is asset data standardisation?
Asset data standardisation is the work of making an organisation's asset records dependable and consistent, so the same asset, component and failure are described the same way at every site. It covers understanding and classifying the asset portfolio, building equipment hierarchies, agreeing a shared ontology of maintainable items, and generating CMMS-ready failure codes.
What does AI readiness mean for asset data?
AI readiness means the underlying asset data is structured, consistent and coded well enough for dashboards, analytics and AI models to produce credible results. Structured data is the prerequisite for those capabilities, not an afterthought once the tools are bought.
What are the nine steps of the framework?
Understand the asset portfolio, classify asset types, build equipment hierarchies, standardise maintainable items, create a reusable knowledge library, generate CMMS-ready failure codes, improve the data feedback loop, steward ownership and governance, and activate digital and AI readiness. They are grouped into five phases: Discover, Structure, Standardise, Govern and Activate.
What is a maintainable item?
A maintainable item is the lowest level in the asset breakdown that is actually maintained, replaced or repaired as a unit. Standardising the names of maintainable items is what makes asset data reusable and CMMS-compatible across sites.
Which industries does the framework apply to?
It is industry agnostic and applies across the heavy asset industries, including mining, oil and gas, rail, smelting and refining, manufacturing and utilities. The method is the same; only the assets differ.
Where should we start?
Start with the data you already have. Tell us where your asset data stands today and we will show you which phase to begin in, whether that is classifying the portfolio, standardising maintainable items, or governing an improvement loop you already run.
Start with the data you already have
Tell us where your asset data stands today and we will show you which phase to start in.