Product Data, Interoperability and Lifecycle Governance

Introduction

Modern mechanical products are developed through a network of computer-aided design (CAD), engineering analysis, manufacturing, quality, purchasing, service and enterprise systems. A three-dimensional model is therefore more than a geometric description. It may be associated with materials, tolerances, manufacturing processes, inspection requirements, software, supplier information, regulatory evidence and change history. Managing these relationships is the purpose of product data management (PDM) and product lifecycle management (PLM).

Poor control of product data can produce incorrect parts, duplicated designs, obsolete drawings, unauthorized changes and costly production delays. Effective governance creates a reliable digital thread: a controlled and traceable chain connecting requirements, design definitions, verification, manufacture, use, maintenance and retirement.

This chapter examines PDM and PLM workflows, bills of materials, versions, configurations, engineering changes, interoperability standards, intellectual property, cybersecurity and data governance.

Learning objectives

By the end of this chapter, the reader should be able to:

1. Distinguish CAD, PDM and PLM responsibilities.
2. Construct and interpret engineering, manufacturing and service bills of materials.
3. Explain versions, revisions, configurations and effectivity.
4. Describe controlled engineering-change workflows.
5. Evaluate STEP, IGES, STL and native CAD exchange.
6. Explain digital threads, data ownership, intellectual property and cybersecurity controls.
7. Recommend governance practices for a manufacturing organization.

1. Product data as an engineering asset

Product data includes geometry, drawings, specifications, materials, tolerances, requirements, test results, software, supplier records, manufacturing instructions and service information. Metadata is equally important. Typical metadata includes part number, title, author, approval status, revision, units, material, applicable product family, release date and retention category.

A useful distinction is between a data object and a document. A document is usually a bounded file such as a drawing or test report. A data object is an item managed with attributes, relationships, permissions and lifecycle states. A CAD file may be both: its contents describe geometry, while the PDM system controls its identity and status.

The main integrity properties of product data are:

• Identity: each controlled item has a unique identifier.
• Traceability: authorship, decisions, approvals and changes are recorded.
• Consistency: related items agree on revision, units and configuration.
• Availability: authorized users can find and use the data when needed.
• Authenticity: users can distinguish approved data from unofficial copies.
• Retention: records remain accessible for the required period.

These properties support safety, quality, cost control and legal defensibility.

2. PDM and PLM architecture

PDM is primarily concerned with controlled product information. It commonly provides vaulting, check-in/check-out, revision control, access control, workflow, search, relationship management and audit trails. PDM is especially concerned with design definition and engineering release.

PLM is broader. It connects product information and processes across the lifecycle, including concept development, requirements, design, analysis, manufacturing, supply, operation, maintenance and disposal. PLM may integrate PDM with enterprise resource planning (ERP), manufacturing execution systems (MES), customer-service platforms, requirements tools and quality systems.

A simplified responsibility model is:

The boundaries vary by organization. The important principle is that interfaces and ownership must be explicit; otherwise the same part number may acquire conflicting meanings in different systems.

Typical controlled workflow

A common workflow is:

1. A user creates or reserves an item identifier.
2. The item is checked out for authorized modification.
3. Design work is performed in CAD.
4. Design reviews and analyses are completed.
5. Related drawings, specifications and BOMs are checked for consistency.
6. The item enters approval workflow.
7. An authorized authority releases it.
8. Manufacturing and downstream systems receive the released definition.
9. Later changes are processed through engineering change control.

“Check-in” should not be confused with approval. A checked-in file may be safely stored but still be in work. Only released data should normally define production, unless a formally controlled deviation exists.

3. Bills of materials and product structure

A bill of materials (BOM) is a structured list of components required to define or produce a product. A BOM may contain item numbers, quantities, units of measure, reference designators, substitutes, effectivity and notes.

The same product can have several legitimate BOM views:

• Engineering BOM (EBOM): organized according to design structure and functional assemblies.
• Manufacturing BOM (MBOM): organized according to production operations, workstations and assembly sequence.
• Service BOM: organized to support maintenance, replacement and field identification.
• As-designed BOM: records the approved design definition.
• As-built BOM: records what was actually assembled in a particular unit.
• As-maintained BOM: records the current configuration after replacements or modifications.

These views must be related without assuming that they are identical. Engineering may model a welded frame as one assembly, while manufacturing may require separate cut, formed, welded and inspected operations. A service organization may identify replaceable modules rather than individual fasteners.

BOM quantity and structure

For a parent item, a component quantity is often represented as the number required per parent assembly. If a product contains several instances of a subassembly, total component demand is obtained by multiplying quantities along each branch of the product tree. For example, if one product contains 2 subassemblies and each subassembly contains 4 brackets, the product requires 2 × 4 = 8 brackets, assuming no scrap factor or alternative configuration.

This simple multiplication is valid only when the structure is unambiguous, quantities are expressed in compatible units and effectivity is the same. It is not valid to multiply across branches that represent mutually exclusive alternatives.

BOM governance problems

Common failures include duplicate part numbers, “phantom” assemblies used inconsistently, obsolete components remaining in released structures, uncontrolled spreadsheet copies and unclear unit conventions. A governed BOM should identify the source of truth, define ownership, enforce mandatory attributes and record the relationship between each parent and child item.

4. Versions, revisions and configurations

A version is a particular state of an item, often created whenever a file is saved or a controlled modification is made. A revision is a formally identified design change that has passed an organizational change process. A version may be temporary; a revision normally has business and technical significance.

A configuration is a selected set of items and revisions that defines a product for a specific purpose. Configuration management answers questions such as:

• Which design was approved on a particular date?
• Which parts were installed in aircraft, vehicles or machines with a given serial number?
• Which components apply to a particular customer, market or regulatory region?
• Which replacement is compatible with an earlier revision?

Configuration rules may use:

• Serial effectivity: applies to specified serial numbers.
• Date effectivity: applies from a defined date.
• Lot or batch effectivity: applies to a production batch.
• Option effectivity: applies when a product option is selected.
• Revision effectivity: applies when a specified revision is used.

A configuration should be reproducible. If two engineers query the same product, revision and effectivity conditions, they should obtain the same controlled structure.

5. Engineering change management

Engineering change management prevents uncontrolled modification of released product definitions. A practical process normally includes an engineering change request (ECR), engineering change order (ECO) and implementation verification.

An ECR describes a problem, opportunity or proposed improvement. It should state the reason, affected products, evidence, urgency, risks and preliminary cost. After analysis, an ECO defines the authorized change, affected items, new revisions, implementation date, disposition of existing stock and required verification.

A robust change process includes:

1. Problem definition: identify the defect, requirement or improvement.
2. Impact analysis: examine geometry, interfaces, strength, function, tooling, inventory, suppliers, software, compliance and service.
3. Alternatives: compare technical and lifecycle consequences.
4. Approval: obtain authorization from relevant engineering, manufacturing, quality, service and commercial authorities.
5. Implementation: update controlled data and communicate the change.
6. Verification and validation: demonstrate that the revised product meets requirements and is fit for intended use.
7. Effectivity and disposition: identify when the change applies and how old stock is treated.
8. Closure: confirm completion and retain objective evidence.

Verification asks whether the design satisfies specified requirements; validation asks whether the resulting product meets user and operational needs. Both may be necessary after a change.

A change classification can help prioritize work:

Organizations must define their own criteria rather than relying only on these labels.

6. Interoperability of product data

Interoperability is the ability of systems, organizations or processes to exchange data and use it correctly. File transfer alone is not interoperability. A receiving system must interpret geometry, units, topology, attributes, structure and design intent sufficiently for its intended purpose.

Native CAD formats

Native formats are created by a particular CAD system. They usually preserve the greatest amount of system-specific information, such as feature history, parametric constraints, assembly relations and design intent. Their disadvantages are dependence on software versions, licensing, translators and vendor-specific structures. Native data should be retained when future editing and design intent are important, but it is not a neutral exchange mechanism.

STEP

STEP is the common name for ISO 10303, a family of standards for the exchange and representation of product data. STEP can represent more than geometric shape, depending on the application protocol and implementation. It may support topology, assemblies, product structure, attributes, tolerances and other product information. AP 203 and AP 214 are historically important for configuration-controlled mechanical design and automotive product data; AP 242 is a modern application protocol intended to improve managed model-based engineering data exchange.

STEP is generally preferred for neutral mechanical CAD exchange when the receiving and sending systems support compatible schemas and translators. Nevertheless, implementation quality varies, and a successful import must be verified.

IGES

IGES, standardized as ANSI Y14.26 and widely used historically, is an older neutral exchange format. It can transfer wireframe, surface and some solid geometry, but it often loses feature history, parametric intent and sophisticated product structure. IGES remains useful for legacy data and certain surface-exchange tasks, but STEP is generally more suitable for modern structured product-data exchange.

STL

STL represents a surface as a triangular tessellation. It is widely used for additive manufacturing, rapid visualization and some inspection workflows. STL normally does not contain native feature history, manufacturing intent, material definition or reliable product structure. It approximates geometry, so chordal deviation, triangle quality, unit interpretation and watertightness matter. STL should not normally be treated as the authoritative definition for precision machining or configuration-controlled design.

Exchange selection

No format guarantees perfect transfer. Organizations should define acceptance tests for geometry, units, tolerances, assembly relationships, metadata and downstream manufacturing results.

7. The digital thread and digital twin

A digital thread is a connected flow of authoritative data and relationships across the product lifecycle. It links requirements to design features, analyses, manufacturing plans, inspection results, field events and changes. The thread is not merely a database or a collection of files; it includes identifiers, semantics, ownership and traceability.

A digital twin is a digital representation of a physical product, system or process that may use operational data to reflect its state or behavior. A CAD model alone is not necessarily a digital twin. A useful twin requires a defined purpose, validated relationships, suitable data refresh, uncertainty awareness and controls over access and interpretation.

Traceability can be represented as a graph: requirements connect to design items; design items connect to analyses and tests; released items connect to BOMs and manufacturing operations; serial products connect to inspection and service records. Missing links identify governance weaknesses and complicate root-cause analysis.

8. Intellectual property and cybersecurity

Product data may contain trade secrets, patented concepts, customer-controlled information, export-controlled technical data and personal information. Ownership and permitted use must be defined in contracts and organizational policies. A supplier may receive a model for manufacturing without receiving rights to reuse it for another customer.

Important controls include:

• data classification and handling labels;
• least-privilege access;
• multi-factor authentication;
• encryption in transit and at rest;
• secure supplier and cloud access;
• immutable or protected audit logs;
• backup, restoration and disaster-recovery testing;
• malware scanning and controlled file exchange;
• segregation of development, test and production environments;
• prompt revocation of access when roles change;
• incident-response and breach-notification procedures.

Cybersecurity must address availability and integrity as well as confidentiality. A malicious alteration to a released machining parameter can be as damaging as data theft. Engineering workstations, post-processors, additive-manufacturing systems and shop-floor networks can all be attack surfaces.

Governance should follow applicable laws and contractual obligations. Depending on the product and location, relevant regimes may include export-control rules, privacy laws, sector-specific safety regulation and standards such as ISO 27001, NIST Cybersecurity Framework guidance, ISO 10303 and ISO 9001 quality-management requirements. Standards do not replace risk assessment or organizational accountability.

9. Data governance framework

Data governance assigns decision rights and responsibilities for product information. A practical framework defines:

• Owners: accountable for the meaning and quality of a data domain.
• Stewards: maintain metadata, quality rules and day-to-day processes.
• Custodians: operate systems, storage and technical controls.
• Users: create, review and consume data according to authorized procedures.
• Approvers: accept technical, quality, regulatory or commercial consequences.

Governance policies should specify naming, numbering, classification, units, metadata, revision rules, retention, archival, access, supplier exchange, approval authority and disposal. Data quality may be assessed through completeness, validity, consistency, uniqueness, timeliness and traceability.

A useful principle is “authoritative once, reused many times.” The organization should avoid retyping released values into disconnected spreadsheets. Interfaces should transfer identifiers and status automatically where practical, while human review remains necessary for safety-critical interpretation.

10. Illustrative case: controlled change in a pump assembly

Background

A manufacturer produces an industrial pump with a cast housing, shaft, impeller, seals and motor. The EBOM is managed in PDM, the MBOM in an MES-linked planning system, and service records in a PLM service module.

Problem

Field reports show premature seal leakage. An engineer proposes changing the seal material and adding a small housing feature. A supplier has already received an exported STEP model of the housing, while production is using an older drawing revision.

Analysis

The team identifies affected items: housing, seal, drawing, EBOM, MBOM, inspection plan, supplier package and service instructions. It discovers that the supplier’s STEP file has correct nominal geometry but lacks the latest tolerance annotation. It also finds finished-goods inventory built to the old revision.

Approach

An ECR records the field evidence. The team performs a failure analysis and tests candidate seals. An ECO then defines new housing and seal revisions, updates the tolerance-controlled drawing, revises the EBOM and MBOM, and specifies serial-number effectivity. Quality updates inspection criteria; purchasing issues a controlled supplier package; service identifies compatible replacement kits. Existing inventory is segregated and dispositioned through an authorized review.

The STEP model is exchanged for geometry, but the released drawing and PDM metadata remain authoritative for tolerances, revision and approval status. The imported model is checked against the released definition before use.

Outcome

The revised assembly is verified by pressure and endurance testing, and validation confirms improved field performance. Each delivered pump is linked to its as-built configuration. The organization can now determine which serial numbers contain the revised seal and can reproduce the design basis for an individual unit.

Learning points

1. A neutral geometry file is not a complete product definition.
2. BOM, drawing, supplier package and service data must change coherently.
3. Effectivity prevents ambiguity between old and new product populations.
4. Configuration traceability supports both safety and customer service.
5. Change control must include inventory and external partners, not only CAD files.

11. Contemporary developments and limitations

Model-based definition increasingly places dimensions, tolerances and notes in a three-dimensional model rather than relying exclusively on a separate drawing. This can reduce inconsistency, but only if viewing tools, semantic tolerancing and manufacturing acceptance processes are mature.

Cloud PLM can improve collaboration and scalability, but it introduces dependency on connectivity, vendor controls, data residency, identity management and service continuity. Artificial intelligence can assist classification, duplicate detection, search, impact analysis and extraction of information from legacy documents. However, AI-generated relationships or recommendations require human validation, especially where safety, regulatory compliance or intellectual property is involved. Training data may also expose confidential designs.

Interoperability remains limited by semantic differences. Two systems may exchange the same shape while disagreeing about units, material state, tolerance meaning, coordinate systems or configuration applicability. Consequently, validation procedures and contractual data definitions remain necessary even when standards-based formats are used.

Best-practice checklist

A mechanically responsible organization should:

1. Assign a unique identity to every controlled product item.
2. Separate work-in-progress, under-review and released states.
3. Define EBOM, MBOM, service and as-built relationships.
4. Record revision, effectivity and approval authority.
5. Use formal impact analysis for engineering changes.
6. Select exchange formats according to the information required.
7. Validate imported geometry, units, metadata and tolerances.
8. Protect intellectual property through classification and least privilege.
9. Maintain auditable access, backups and recovery procedures.
10. Treat cybersecurity as a product-integrity and safety concern.
11. Connect requirements, design, manufacturing, inspection and service data.
12. Use AI and automation as controlled aids rather than unquestioned authorities.

Chapter Summary

Product data is a controlled engineering asset that includes geometry, structure, specifications, metadata, approvals and lifecycle evidence. PDM controls engineering information and workflows, while PLM connects product information across design, manufacture, operation and retirement. BOMs must be interpreted according to their purpose, because engineering, manufacturing, service, as-built and as-maintained structures are related but not identical.

Versions represent states of data, revisions represent controlled design changes, and configurations combine approved items under effectivity rules. Engineering change management requires impact analysis, authorization, implementation, verification, validation and disposition of existing products or stock.

Native CAD data preserves design intent but is system-dependent. STEP is a major neutral standard for structured product data; IGES is valuable mainly for legacy exchange; STL is a tessellated geometry representation suited to mesh-based applications rather than authoritative precision product definition. Digital threads connect lifecycle evidence, while digital twins represent physical products or processes for defined operational purposes.

Intellectual-property protection, cybersecurity, access control, auditability, retention and data quality are essential elements of lifecycle governance. Effective organizations establish ownership, stewardship, authoritative sources, controlled interfaces and reproducible configurations.

Key Terms

As-built configuration: The recorded configuration actually manufactured for a particular product.

As-maintained configuration: The current configuration after service replacements or modifications.

Bill of materials (BOM): A structured list of components and quantities forming a product or assembly.

Configuration management: Control of product identity, structure, status, revision and change throughout the lifecycle.

Digital thread: Connected, traceable lifecycle data linking requirements, design, manufacture, operation and service.

Digital twin: A digital representation of a physical product, system or process maintained for a defined purpose.

Effectivity: A rule specifying when, where or to which product population a definition applies.

Engineering change order (ECO): An authorized instruction defining and implementing a product change.

Engineering change request (ECR): A proposal or problem report initiating investigation of a possible change.

Engineering BOM (EBOM): A BOM organized according to the engineering design structure.

IGES: A historical neutral format for exchanging CAD geometry, often with limited design-intent retention.

Metadata: Data describing an information object, such as its identifier, revision, author or status.

Native CAD format: A proprietary format produced by a particular CAD application.

PDM: Product data management for controlled storage, relationships, revisions, permissions and workflows.

PLM: Product lifecycle management across the complete product lifecycle.

Revision: A formally controlled state representing an approved design change.

STL: A triangular surface-tessellation format commonly used in additive manufacturing.

STEP: ISO 10303, a family of standards for product-data representation and exchange.

Version: A particular saved or controlled state of an information item.

Review Questions

Short Answer Questions

1. What is the principal difference between PDM and PLM?
2. Why are EBOM and MBOM often different?
3. Distinguish a version from a revision.
4. What information does effectivity provide?
5. Why is STL usually unsuitable as the sole authoritative product definition?

Long Answer Questions

1. Explain a complete engineering-change workflow from problem identification to closure.
2. Compare native CAD, STEP, IGES and STL in terms of geometry, design intent and typical use.
3. Describe how a digital thread supports configuration traceability.
4. Discuss the roles of data owners, stewards, custodians and users in product-data governance.
5. Explain why cybersecurity is a product-integrity issue in a connected manufacturing environment.

Analytical / Application Questions

1. A supplier receives a STEP model, while production uses a drawing with a different revision. Analyze the risks and propose a controlled resolution.
2. Design a BOM and effectivity strategy for a machine offered with two mutually exclusive motor options and a later safety upgrade.
3. An organization wants to use generative AI to identify duplicate CAD parts and recommend engineering-change impacts. Evaluate the benefits, risks and governance controls required.

Multiple Choice Questions with Answers

Question 1

Which system is primarily responsible for controlled engineering files, revisions and approval workflows?

A. PDM

B. ERP

C. MES

D. QMS

Correct Answer: A. PDM

Question 2

Which BOM is normally organized around production operations and workstations?

A. Engineering BOM

B. Manufacturing BOM

C. Service BOM

D. As-maintained BOM

Correct Answer: B. Manufacturing BOM

Question 3

What does serial-number effectivity specify?

A. The CAD software version used

B. The material supplier identity

C. The product serial numbers to which a definition applies

D. The number of triangles in an STL file

Correct Answer: C. The product serial numbers to which a definition applies

Question 4

Which format is an ISO 10303 family of standards?

A. IGES

B. STL

C. STEP

D. DXF

Correct Answer: C. STEP

Question 5

What is a principal limitation of STL?

A. It cannot represent a tessellated surface

B. It normally lacks feature history and design intent

C. It is restricted to electrical schematics

D. It cannot be used for additive manufacturing

Correct Answer: B. It normally lacks feature history and design intent

Question 6

What is the main purpose of an engineering change order?

A. To authorize and define implementation of a product change

B. To compress a CAD file

C. To replace all service records

D. To create an employee account

Correct Answer: A. To authorize and define implementation of a product change

Question 7

Which statement best describes a digital thread?

A. A physical cable connecting machines

B. A single native CAD file

C. A connected and traceable flow of lifecycle data

D. A triangular mesh used in inspection

Correct Answer: C. A connected and traceable flow of lifecycle data

Question 8

Which cybersecurity principle grants users only the access needed for their roles?

A. Least privilege

B. Open access

C. Uncontrolled replication

D. Anonymous sharing

Correct Answer: A. Least privilege

Question 9

A product has 3 subassemblies, and each contains 5 identical brackets. Ignoring alternatives and scrap, how many brackets are required?

A. 8

B. 15

C. 20

D. 35

Correct Answer: B. 15

Question 10

Which statement about a checked-in CAD file is correct?

A. It is automatically approved for production

B. It is safely stored but may still be work-in-progress

C. It must be an STL file

D. It cannot have metadata

Correct Answer: B. It is safely stored but may still be work-in-progress