Competition Law and Predictive Maintenance Platform Competition

1. Introduction

Predictive maintenance platforms use sensors, Internet of Things (IoT) systems, machine-learning models, cloud computing and industrial data to predict when equipment is likely to fail and recommend maintenance before a breakdown occurs.

They are increasingly important in:

manufacturing;

aviation;

railways;

automobiles;

shipping;

power generation;

oil and gas;

mining;

healthcare equipment;

industrial robotics;

renewable-energy installations;

smart infrastructure.

From a competition-law perspective, predictive maintenance creates a distinctive combination of data concentration, software dependency, aftermarket power, interoperability issues, vertical integration, intellectual property rights and ecosystem effects.

A manufacturer may supply a machine, control its diagnostic software, collect operational data, provide cloud-based analytics and sell maintenance services. If competitors require access to that information or software to provide independent maintenance, competition may be affected.

The central question is therefore:

When does control over predictive-maintenance technology, machine data, diagnostic information or connected-service infrastructure give an undertaking the ability or incentive to restrict competition in maintenance and related markets?

Predictive maintenance itself is not anticompetitive. Competition-law concerns arise from the way a platform uses its technological, informational or contractual position.

2. What Is a Predictive Maintenance Platform?

A predictive-maintenance platform generally performs several functions:

Data collection — sensors collect information concerning machines.

Data transmission — information is transferred through IoT networks.

Data storage — information is stored in cloud or proprietary databases.

Analytics — algorithms analyse equipment behaviour.

Prediction — the system predicts potential failures.

Recommendation — the platform recommends maintenance actions.

Service integration — maintenance providers receive or implement recommendations.

The ecosystem may therefore involve:

Machine → Sensor → Data → Cloud → Algorithm → Prediction → Maintenance → Replacement Parts

Competition problems can occur at any of these layers.

3. Relevant Markets

Predictive maintenance can produce several separate but interconnected markets.

A. Equipment market

Examples include:

aircraft engines;

turbines;

industrial machinery;

locomotives;

medical equipment.

B. Diagnostic software market

This includes software used to identify:

faults;

wear;

performance deterioration;

component failure.

C. Predictive analytics market

This involves algorithms and AI systems that forecast equipment failure.

D. Maintenance services market

Independent maintenance providers may compete with the equipment manufacturer.

E. Spare-parts market

The manufacturer may produce proprietary replacement parts.

F. Data-access market

Industrial data itself may become an important competitive input.

G. Connected-service ecosystem

The manufacturer may combine:

equipment;

software;

cloud services;

maintenance;

spare parts;

financing.

The relevant market therefore cannot automatically be assumed to be the market for "predictive maintenance." Several markets may need to be analysed separately or as interconnected markets.

4. Aftermarket Competition

Predictive maintenance raises classic aftermarket questions.

Consider a manufacturer that sells an industrial machine for ₹10 crore.

After installation, customers need:

diagnostics;

software;

spare parts;

servicing;

updates;

technical information.

If the manufacturer controls those services, the initial equipment market may be competitive while the aftermarket becomes highly concentrated.

Competition authorities may therefore examine:

switching costs;

installed base;

contractual restrictions;

availability of independent maintenance;

access to diagnostic information;

spare-parts compatibility.

5. Data as a Competitive Input

Predictive maintenance depends heavily on operational data.

Examples include:

temperature;

vibration;

pressure;

energy consumption;

component degradation;

operating cycles;

failure history.

A manufacturer with millions of connected machines may possess a dataset that competitors cannot easily reproduce.

This creates a potential data-based entry barrier.

However, possession of valuable data does not automatically create dominance or a duty to provide access. The legal analysis must establish the relevant market and the applicable competition-law requirements.

6. Data Lock-In

Customers may become dependent upon a manufacturer's historical maintenance database.

When changing suppliers, they may lose:

historical maintenance records;

machine-learning models;

diagnostic history;

failure predictions;

component histories.

This can make switching costly.

Competition concerns become stronger where contractual or technical restrictions prevent customers from transferring their data to competing maintenance providers.

7. Refusal to Provide Diagnostic Data

A manufacturer may refuse to provide independent repairers with:

diagnostic codes;

technical manuals;

machine data;

API access;

software interfaces;

calibration information.

This may restrict independent maintenance competition.

However, a refusal-to-deal theory normally requires careful analysis of dominance, indispensability, elimination of effective competition and other applicable conditions.

8. Essential-Facility Considerations

The doctrine developed in cases such as Bronner and IMS Health may become relevant where competitors claim that access to a particular technical system or information resource is indispensable.

A predictive-maintenance platform could theoretically become an important facility if:

customers are locked into the system;

independent providers cannot reasonably reproduce the relevant information;

access is technically controlled by the manufacturer;

denial of access eliminates effective downstream competition.

But competition law does not automatically transform every proprietary database or software system into an essential facility.

9. Interoperability

Interoperability is central to predictive maintenance.

Independent service providers may need:

APIs;

machine-readable data;

communication protocols;

diagnostic interfaces;

authentication systems.

A dominant manufacturer could potentially restrict interoperability to protect its maintenance business.

Relevant competition-law questions include:

Is the interface technically indispensable?

Can rivals develop alternative methods?

Is access commercially reasonable?

Is the restriction objectively justified?

Does the restriction eliminate effective competition?

10. Bundling and Tying

Predictive maintenance can be tied to equipment sales.

For example:

"Customers purchasing this machine must use our predictive-maintenance subscription."

Potentially separate products may include:

machine;

diagnostic software;

cloud service;

maintenance;

spare parts.

If a dominant undertaking conditions access to one product upon purchasing another, Article 102 TFEU and Section 4 of the Indian Competition Act may become relevant.

11. Self-Preferencing

A predictive-maintenance platform may operate both:

as the data/analytics infrastructure provider; and

as a maintenance service provider.

It may then control the ranking of maintenance recommendations.

For example, an algorithm could recommend:

Manufacturer's maintenance service — first
Independent provider — second

If the platform systematically favours its own downstream service without adequate justification, competition concerns may arise.

The legal assessment would depend on dominance, conduct, foreclosure effects and the relevant market structure.

12. Predictive Maintenance and Spare Parts

Predictive maintenance may generate information concerning the exact part that is likely to fail.

The platform could therefore control access to:

component specifications;

failure codes;

replacement recommendations;

compatible parts.

If the equipment manufacturer also sells spare parts, it may have incentives to favour its own parts.

This creates a potential vertical foreclosure issue:

Equipment → Diagnostics → Maintenance → Spare Parts

Control over one layer may be leveraged into another.

13. Algorithmic Discrimination

A predictive-maintenance platform may apply different commercial conditions based on predicted customer behaviour.

For example, it could predict that:

a large customer is unlikely to switch;

a small customer is highly dependent;

an independent maintenance provider is growing rapidly.

The platform could then offer different:

prices;

access terms;

software functionality;

data access;

service levels.

Personalisation is not automatically unlawful, but discriminatory conduct by a dominant undertaking may become relevant under abuse-of-dominance rules.

14. Competition Between OEMs and Independent Maintenance Providers

Original equipment manufacturers (OEMs) often possess advantages because they:

designed the equipment;

know its architecture;

control diagnostic software;

possess historical data;

manufacture spare parts.

Independent maintenance firms may nevertheless compete through:

lower prices;

specialist expertise;

multi-brand servicing;

faster response;

innovative analytics.

Competition law can become relevant if the OEM uses its upstream position to systematically prevent independent firms from competing downstream.

15. Key Case Laws

15.1 Bronner v Mediaprint

Case: C-7/97

The Court of Justice established important principles governing refusal by a dominant undertaking to provide access to infrastructure.

The Court applied a demanding test involving indispensability and elimination of effective competition.

Relevance to predictive maintenance

If independent maintenance providers request access to:

proprietary diagnostic infrastructure;

machine-data systems;

essential interfaces;

Bronner provides an important analytical starting point.

Not every useful database or interface is legally indispensable.

15.2 IMS Health v Commission

Case: C-418/01

The case concerned access to a copyrighted database and the circumstances in which refusal to license intellectual property may constitute an abuse.

The Court identified exceptional circumstances relevant to compulsory access.

Predictive-maintenance relevance

Predictive maintenance can involve proprietary:

datasets;

software;

databases;

diagnostic structures.

IMS Health demonstrates the need to balance intellectual-property protection with exceptional circumstances supporting compulsory access.

15.3 Microsoft Corp. v Commission

Case: T-201/04

Microsoft involved interoperability information and the relationship between dominance, technological interfaces and downstream competition.

The General Court upheld important aspects of the Commission's findings concerning Microsoft's conduct.

Relevance

Predictive-maintenance platforms frequently depend upon interoperability.

A manufacturer controlling machine interfaces could potentially restrict competitors' ability to interact with equipment.

Microsoft therefore provides an important framework for analysing technical interoperability and foreclosure.

15.4 Volvo v Veng

Case: 238/87

The case concerned intellectual property rights and the supply of replacement parts.

The Court recognised that exercising an intellectual-property right does not automatically constitute abuse, while also recognising circumstances in which refusal to supply may raise Article 102 issues.

Predictive-maintenance significance

The case is particularly relevant to the relationship between:

proprietary equipment;

spare parts;

intellectual property;

downstream repair markets.

Predictive-maintenance systems can strengthen OEM control over spare-parts ecosystems.

15.5 Magill

Joined Cases: C-241/91 P and C-242/91 P

Magill established important principles concerning exceptional circumstances in which refusal to license intellectual property may constitute abuse.

Relevance

Predictive-maintenance platforms may contain proprietary information necessary for competing maintenance providers.

Magill illustrates that intellectual-property protection and competition law must sometimes be balanced where refusal to provide access has serious downstream competitive effects.

15.6 Commercial Solvents

Joined Cases: 6/73 and 7/73

Commercial Solvents concerned a dominant undertaking's refusal to supply an important input to downstream competitors.

The Court treated the use of upstream dominance to eliminate downstream competition as potentially abusive.

Predictive-maintenance application

The principle can be applied conceptually where a manufacturer controls an upstream input such as:

diagnostic information;

machine data;

software access;

critical components.

If the upstream input is necessary for downstream competition, foreclosure concerns may arise.

15.7 Slovak Telekom

Cases: C-165/19 P and related proceedings

The case concerned access restrictions and exclusionary effects in telecommunications.

The Court examined the relationship between dominance, access conditions and foreclosure.

Predictive-maintenance relevance

Although telecommunications differs from industrial maintenance, the principles concerning access restrictions and exclusionary effects are relevant to connected maintenance ecosystems.

A predictive-maintenance provider may similarly control an upstream infrastructure layer upon which downstream service providers depend.

15.8 Deutsche Telekom

Case: C-280/08 P

The case concerned margin squeeze and exclusionary effects in telecommunications markets.

Relevance

Predictive-maintenance ecosystems can create vertically integrated structures:

Equipment → Data → Platform → Maintenance

If the same undertaking controls upstream access and competes downstream, pricing structures can potentially create margin-squeeze concerns where the legal requirements are satisfied.

15.9 Microsoft Mobile / Microsoft Interoperability Principles

The broader Microsoft jurisprudence demonstrates that competition law may intervene where a dominant undertaking uses technological control over an important ecosystem component to disadvantage competing products or services.

For predictive maintenance, this can arise through:

proprietary interfaces;

software compatibility;

access protocols;

authentication;

cloud integration.

15.10 Google Shopping

Case: T-612/17

The case concerned Google's treatment of its comparison-shopping service in search results.

Relevance

The case provides an important digital-platform analogy for predictive maintenance.

A platform that controls:

data;

rankings;

recommendations;

access;

and simultaneously competes downstream may create concerns where those functions are used to favour its own service and restrict rivals.

16. Predictive Maintenance and Merger Control

Merger control can become important where an OEM acquires a predictive-maintenance company.

The transaction may combine:

machine data;

AI models;

maintenance customers;

spare-parts information;

equipment installed bases.

Authorities may examine whether the merger could:

increase data concentration;

eliminate an emerging competitor;

reduce innovation;

foreclose independent maintenance providers;

increase switching costs;

strengthen vertical integration.

The target's current revenue may not fully capture its competitive significance if its technology has substantial future potential.

17. Vertical Foreclosure

Predictive maintenance creates several possible vertical relationships:

Equipment Manufacturer
↓
Diagnostic Software
↓
Cloud Platform
↓
Predictive Analytics
↓
Maintenance Provider
↓
Spare Parts

An integrated undertaking may have both the ability and incentive to restrict competitors.

Competition authorities would typically need to examine:

market power;

ability to foreclose;

incentive;

actual or likely effects;

efficiencies;

availability of alternatives.

18. Platform Neutrality

Where a predictive-maintenance platform serves multiple maintenance companies, neutrality becomes important.

Potential concerns include:

ranking manipulation;

discriminatory API access;

discriminatory subscription terms;

preferential data access;

differential technical functionality;

preferential recommendations.

The platform may effectively function as a market intermediary.

Its role therefore resembles other digital platforms where the operator simultaneously governs and participates in the market.

19. Competition and Customer Data Ownership

One of the most important contractual issues concerns who controls machine-generated data.

Contracts may state that:

the OEM owns all data;

the customer owns the data;

the platform receives an exclusive licence;

data can be used only within the OEM ecosystem.

Competition implications may arise where restrictive contractual terms make it difficult for customers to use independent maintenance providers.

The competition analysis should distinguish:

ownership

from

access

from

portability

from

commercial reuse.

They are legally distinct questions.

20. Security Justifications

Manufacturers may legitimately restrict access to diagnostic systems because of:

cybersecurity;

safety;

product liability;

protection against malicious modification;

system integrity.

Competition law should therefore consider whether restrictions are objectively justified.

A blanket refusal based merely on "security" may require scrutiny if less restrictive technical solutions exist.

Possible alternatives include:

read-only APIs;

certification systems;

authenticated access;

controlled diagnostic interfaces;

audit trails.

21. Predictive Maintenance and Right-to-Repair

Predictive maintenance is closely connected to the broader right-to-repair debate.

Independent repairers may require:

technical manuals;

diagnostic information;

software tools;

spare parts;

machine data.

Competition law can complement sector-specific repair legislation, but the two areas should not be conflated.

A conduct may be:

regulated by right-to-repair legislation;

relevant to competition law;

both;

or neither.

22. Indian Competition-Law Perspective

The Indian Competition Act, 2002 provides several possible analytical routes.

Section 3

Relevant where manufacturers or service providers coordinate:

maintenance prices;

spare-parts allocation;

repair territories;

service restrictions.

Section 4

Potential abuse-of-dominance issues include:

denial of market access;

discriminatory conditions;

tying;

leveraging;

exclusionary conduct.

Sections 5 and 6

Relevant mergers may involve combinations of:

equipment manufacturers;

industrial software providers;

IoT platforms;

predictive-analytics companies.

Digital-market dimension

The Competition Commission of India can examine market power beyond simple physical-product market shares where digital infrastructure, data, network effects and switching costs affect competitive conditions.

23. Competition-Risk Matrix

These indicators do not by themselves establish an infringement.

24. How Competition Authorities Could Analyse a Predictive-Maintenance Case

A structured analysis could proceed as follows.

Step 1 — Define the relevant market

Determine whether competition occurs in:

equipment;

maintenance;

diagnostics;

predictive analytics;

spare parts;

connected services.

Step 2 — Establish market power

Consider:

installed base;

market share;

network effects;

switching costs;

data advantages;

technological barriers.

Step 3 — Identify the controlled input

Ask whether the undertaking controls:

data;

software;

APIs;

diagnostic tools;

spare parts;

cloud infrastructure.

Step 4 — Identify exclusionary conduct

Determine whether it:

refuses access;

degrades interoperability;

ties products;

discriminates;

self-preferences;

imposes exclusivity.

Step 5 — Examine competitive effects

Assess:

foreclosure;

price effects;

innovation;

service quality;

consumer choice;

entry;

independent repair competition.

Step 6 — Consider efficiencies

Examine:

safety;

cybersecurity;

reliability;

reduced downtime;

lower maintenance costs;

technological innovation.

25. Important Distinction: Prediction Is Not Market Power

An undertaking may possess excellent predictive technology without being dominant.

Similarly:

Large amounts of machine data do not automatically establish an essential facility.

And:

A proprietary algorithm does not automatically create an abuse of dominance.

Competition law requires the relevant statutory elements to be established.

This is particularly important because predictive maintenance can produce substantial consumer and industrial benefits.

26. Economic Benefits of Predictive Maintenance

Competition analysis should account for legitimate efficiencies such as:

reduced equipment downtime;

reduced maintenance costs;

improved safety;

lower energy consumption;

longer equipment life;

reduced waste;

better spare-parts planning;

improved industrial productivity.

A competition authority must therefore distinguish between efficient ecosystem integration and exclusionary ecosystem control.

27. Core Competition-Law Questions

The most important questions are:

Who controls the machine data?

Can customers transfer their data?

Can independent repairers access necessary diagnostic information?

Can rival predictive-maintenance software interoperate?

Does the OEM compete with independent maintenance providers?

Does the OEM favour its own maintenance services?

Are spare parts restricted?

Are maintenance services tied to equipment purchases?

Does the platform impose exclusivity?

Does a merger combine critical datasets and predictive technologies?

28. Conclusion

Predictive-maintenance platforms illustrate how competition law is increasingly moving beyond conventional product markets toward data-driven industrial ecosystems.

The most significant competition concerns arise when a firm simultaneously controls:

equipment + machine data + diagnostic software + predictive analytics + maintenance + spare parts.

Such vertical integration can generate efficiencies, but it can also create opportunities for:

input foreclosure;

aftermarket exploitation;

interoperability restrictions;

tying;

discriminatory access;

self-preferencing;

data lock-in;

exclusion of independent maintenance providers;

strategic acquisitions of emerging competitors.

The cases of Bronner, IMS Health, Microsoft, Volvo v Veng, Magill, Commercial Solvents, Slovak Telekom, Deutsche Telekom and Google Shopping provide useful doctrinal foundations for analysing these issues.

The central competition-law principle is therefore that predictive-maintenance technology should not be treated as anticompetitive merely because it is proprietary or technologically sophisticated. The decisive questions concern market power, control over competitively significant inputs, exclusionary conduct, effects on competition, and legitimate technological or efficiency justifications.

Case-Law Summary