India’s First Olfactory Trade mark: Reading the Sumitomo Order with Critical Appreciation

An analysis of India’s first olfactory trade mark registration for Sumitomo Rubber Industries’ rose-like fragrance applied to tyres, examining the statutory framework, graphical representation, distinctiveness, international approaches, and the unresolved challenges surrounding enforcement of smell marks in India.

Introduction

Every trade mark system, at some point, confronts a mark it was not designed to accommodate. For India, that confrontation arrived on 21 November 2025, when the Controller General of Patents, Designs and Trade Marks passed an order accepting Application No. 5860303 “Floral Fragrance/Smell Reminiscent of Roses as Applied to Tyres” filed by Sumitomo Rubber Industries Ltd., Japan (hereinafter “the order”).¹ It is India’s first olfactory trade mark. Until that order, not one smell mark had been registered or accepted in this country.

The order deserves neither uncritical celebration nor reflexive scepticism. It is a serious, carefully reasoned response to questions that have troubled trade mark registries across the world for three decades. It draws on international jurisprudence, commissions scientific expertise, and applies the statute purposively. At the same time, it leaves open questions on representation standards, on distinctiveness, and on enforcement that the first order in any new area of law inevitably cannot answer alone. This article analyses the order across three dimensions: the statutory framework, the graphical representation solution, and the distinctiveness finding. It then identifies what remains unresolved and what that means for those who will apply this precedent going forward.

The statutory question: Silence is not prohibition

The Trade Marks Act, 1999 (1999 Act) defines “trade mark” in Section 2(1)(zb) as a mark “capable of being represented graphically” and “capable of distinguishing the goods or services of one person from those of others”. The definition of “mark” in Section 2(1)(m) includes devices, brands, words, numerals, shapes, packaging, and combinations of colours followed by the phrase “or any combination thereof”.

Smell marks appear nowhere in this list. But the list is introduced by the word “includes” a term of settled legal significance. The Supreme Court has consistently held that an inclusive definition is expansive by design, intended to encompass categories beyond those explicitly enumerated.² The legislature’s use of “includes” in Section 2(1)(m) was not inadvertent; it reflects an awareness that commerce evolves and that marks may take forms not anticipated at the time of drafting.

The 1999 Act also does not, anywhere, prohibit smell marks. This is significant. The Controller General correctly observed that legislative silence is not legislative prohibition, and that the absence of an explicit mention of olfactory marks cannot, by itself, be a ground for refusal.³ The relevant question is not whether the Act anticipated smell marks but whether the Act’s requirements, properly construed, can be satisfied by a smell. That is the real inquiry, and it breaks into two parts: graphical representation and distinctiveness.

One further point on the statutory architecture deserves mention. The Trade and Merchandise Marks Act, 1958, the predecessor legislation imposed no graphical representation requirement at all.⁴ The 1999 Act introduced this requirement deliberately, to ensure that the register is certain, accessible, and objectively defined.⁵ This legislative intent must inform the standard applied to any representation, including one for a smell. The requirement was introduced to protect the public, not to create an arbitrary exclusion of entire sensory categories.

Graphical representation: An innovative solution and its limits

1. The Sieckmann standard

The international benchmark on this question is the European Court of Justice’s (ECJ) judgment in Sieckmann v. Deutsches Patent- und Markenamt⁶. The applicant sought to register an olfactory mark for methyl cinnamate, supported by a chemical formula, a verbal description, and an odour sample. The ECJ rejected all three, stating that the chemical formula describes a substance, not an experience; the verbal description lacks precision and objectivity; and the odour sample is neither stable over time nor graphically representable. The court held that a valid graphical representation of any mark must be clear, precise, self-contained, easily accessible, intelligible, durable, and objective, all seven criteria must be satisfied simultaneously.⁷

The practical consequence of Sieckmann in Europe was to close the door on olfactory marks. No smell has since been registered in the European Union (EU)⁸, not because the law forbids it, but because no available technology could meet all seven criteria at once. The Sieckmann criteria are not binding on Indian Courts or the Registry. But they represent the most rigorously reasoned international standard available, and the order under analysis consciously applies them as the evaluative framework.⁹

2. The seven-dimensional vector: A genuine advance

The graphical representation accepted by the Registry was prepared by Prof. Pritish Varadwaj, Prof. Neetesh Purohit, and Dr Suneet Yadav of the Indian Institute of Information Technology (IIIT), Allahabad.¹⁰ Using supervised machine learning, they identified the top five volatile organic compounds constituting a rose-like smell and, by algebraic averaging of their respective vectors, mapped the smell as a point in seven-dimensional space, each dimension corresponding to one of seven fundamental smell categories: floral, fruity, woody, nutty, pungent, sweet, and minty. The output is a radar chart: A two-dimensional visual projection of this seven-dimensional mapping.

This is genuinely innovative. Where Sieckmann rejected a chemical formula as representing the substance rather than the smell, the IIIT, Allahabad methodology works at the level of olfactory perception, it maps how the smell is experienced, not merely what it is composed of. The Controller General accepted this as satisfying the graphical representation requirement, finding that it defines the constituent elements of the smell with scientific precision and enables competent authorities to determine the subject-matter of protection.¹¹

That finding is defensible. But leaves concerns.

(a) The accessibility problem

The Sieckmann standard requires a representation to be “easily accessible”. This requirement protects the public notice function of the trade mark register, a competitor planning, a product, a lawyer advising on clearance, or a court adjudicating infringement must be able to look at the register and understand what is protected. A seven-dimensional olfactory vector produced by machine learning and algebraic averaging of chemical compound data is not, without expert assistance, interpretable by a non-specialist. The radar chart conveys that the smell has certain proportions across seven categories; it does not convey what “the smell” smells like to someone reading the register. Whether this satisfies the “easily accessible” and “intelligible” limbs of the Sieckmann standard is a question the order does not squarely engage.

One possible remedy would be to anchor the radar chart to a standardised verbal descriptor which may be a controlled, objectively defined olfactory reference point so that a non-specialist reader can cross-reference the chart against a publicly available sensory lexicon. The Registry would be well-placed to develop such a lexicon through consultation with the Bureau of Indian Standards or a designated scientific body.

(b) The resolution problem

The seven fundamental smell categories are broad. A rose and a jasmine may produce different radar charts. But a rose and a similar floral fragrance, let us say “a peony” or “a geranium” may produce charts that are indistinguishable within the model’s resolution. If two perceptibly distinct smells produce identical graphical representations, the representation fails its core purpose: Defining the boundaries of protection. As more olfactory applications are filed, this limitation will matter.¹²

The Registry should, through examination guidelines, prescribe a minimum resolution threshold expressed in terms of the model’s discriminatory sensitivity below which a representation will not be accepted as sufficiently precise.

(c) The reproducibility question

A graphical representation must be durable and objective. Whether the same methodology, applied by a different team of scientists to the same rose smell, produces an identical radar chart is not established in the order. Machine learning outputs can vary with training data and algorithmic parameters. If the representation is not reproducible by an independent operator, its objectivity is incomplete.

The Registry should require, as a condition of acceptance, that the applicant deposit the full algorithmic parameters and training dataset with the Registry, so that an independent scientific body can verify reproducibility.

These are not reasons to refuse the application. They are reasons to refine the methodology in subsequent guidelines before it is treated as a settled standard.

Distinctiveness is the order’s strongest ground and one overlooked tension

Arbitrariness as the basis for inherent distinctiveness

Section 9(1), 1999 Act bars registration of marks devoid of distinctive character or marks that designate the quality, kind, or characteristics of goods.¹³ The Controller General held that the rose fragrance is inherently distinctive for tyres because it bears no natural, functional, or descriptive relationship to rubber products.¹⁴ A rose scent is not produced by the manufacturing process; it does not describe a characteristic of the tyre; it serves no functional purpose in the tyre’s performance. Applied to tyres, it is, in trade mark parlance, arbitrary, and arbitrary marks are inherently distinctive.

This reasoning is sound, and it is the order’s most persuasive passage. The analogy is straightforward: “Apple” applied to computers is arbitrary for the same reason. A consumer standing near a highway expects rubber fumes. Roses are so far removed from that expectation that any consumer who perceives a rose scent from a passing vehicle will register it as a deliberate brand signal, not a natural product characteristic. That is the essence of a source-identifying mark.

The comparison with the world’s first registered smell mark Celia Clarke, In re (Clarke’s Osewez)¹⁵ is instructive. The United States (US) Trademark Trial and Appeal Board (TTAB) registered “a high impact, fresh, floral fragrance reminiscent of Plumeria blossoms” for sewing thread and embroidery yarn because the scent was non-functional, was added by the applicant as a distinguishing feature, and had come to be recognised by consumers as identifying that applicant’s goods. The logic of non-functionality plus source identification applies equally here.

It is worth noting, however, that the US takes the position that smell marks cannot be inherently distinctive and requires proof of acquired distinctiveness through use.¹⁶ The Indian approach of accepting inherent distinctiveness on the basis of arbitrariness is therefore more permissive than the US standard, and this divergence should be acknowledged in any comparative analysis of the order’s global significance.

The “proposed to be used” problem

Here the order is less settled, and intellectual honesty requires the tension to be named. The Sumitomo application was filed on a ”proposed-to-be-used” basis. There is no evidence that rose-scented tyres have been sold in India, marketed in India, or encountered by a single Indian consumer. The distinctiveness finding rests entirely on abstract arbitrariness and the theoretical non-relationship between roses and tyres without any foundation in actual market perception.

Trade mark law ultimately protects consumer associations in the marketplace, not theoretical combinations on paper. Distinctiveness in the abstract is a workable ground for registration under the 1999 Act, the statute permits proposed-to-be-used applications. But in the context of a non-conventional mark that no Indian consumer has ever encountered, the question of whether the mark functions as a source identifier in India is deferred rather than answered. The United Kingdom (UK) registration (UK Trade Mark Reg. No. UK00002001416, applied 1994, registered 1996)¹⁷, the world’s first smell mark was cited in submissions as evidence of international recognition. It is relevant as jurisprudential precedent. It does not establish distinctiveness in the Indian market, and international reputation, absent a well-known mark finding under Section 11, 1999 Act¹⁸ does not substitute for it.

It is worth noting that Sumitomo could, in future proceedings, seek a well-known mark declaration under Section 11, 1999 Act read with Rule 124, Trade Marks Rules, 2017. Such a declaration, if granted on the basis of the mark’s international recognition and long-standing use in the UK and other jurisdictions, would provide a stronger distinctiveness foundation in India than the present registration alone.

This concern does not invalidate the registration. It does mean that the distinctiveness foundation, though legally sufficient, is relatively thin and will be tested seriously if infringement proceedings are ever brought.

The role of the Amicus Curiae

The Registry appointed Shri Pravin Anand as Amicus Curiae, acknowledging the novelty of the legal and scientific questions involved.¹⁹ The appointment was appropriate. This application traversed trade mark law, sensory science, and machine learning methodology, a combination no examining officer should navigate without neutral expert assistance. His contributions along with comparative jurisprudential analysis across five jurisdictions, the scientific framework, and the coordination of the IIIT, Allahabad representation were genuinely material to the outcome. What deserves brief reflection is that the graphical representation itself was first placed on record by the amicus and subsequently adopted by the applicant.²⁰ In unopposed examination proceedings, as here, this creates no procedural difficulty. In a contested opposition proceeding with a third-party opponent, an arrangement where the court’s neutral effectively supplies the key instrument resolving the applicant’s principal objection would be procedurally more exposed and may invite challenge on grounds of procedural fairness under the Trade Marks Rules, 2017. As smell mark applications multiply and they will, following this order the Registry will need institutionalised examination guidelines for olfactory marks, not case-by-case appointments of Senior Counsel, however capable.

The development of such guidelines deserves to be treated as a matter of institutional priority. The Registry should consider constituting a Standing Expert Committee comprising trade mark practitioners, sensory scientists, and information technology specialists to draft examinations standards for non-conventional marks, with olfactory marks as the first subject of attention.

The enforcement gaps

The order directs advertisement under Section 20, 1999 Act.²¹ Full registration follows the opposition period. But assuming registration, the more pressing question is: What happens next?

Section 29, 1999 Act²² empowers the owner of a registered trade mark to sue for infringement where an identical or deceptively similar mark is used without authorisation. For a word or device mark, deceptive similarity is assessed visually and phonetically. For a smell mark, a court must compare two smells. There is no established procedure in Indian civil litigation for olfactory evidence. There is no standard for how a scent sample is produced, authenticated, and preserved for court. There is no precedent for whether trained sensory panels, gas chromatography-mass spectrometry (GC-MS) analysis²³ would constitute documentary evidence, subject to authentication and expert certification, or machine-learning comparison would raise questions of admissibility of algorithmically generated evidence. This is a question that Indian Courts have not yet addressed in any reported decision of olfactory vectors. The Registry and the legislature should consider which of these mechanisms is to be treated as the primary evidentiary standard, and rules amendments or practice directions should be issued accordingly.

A registered right without a reliable remedy is, in practice, a paper right. The order creates the right. The enforcement architecture whether built through Registry guidelines, rules amendments, or the first infringement court’s judgment remains entirely to be constructed.

The global context

India arrives at olfactory trade mark protection three decades after the first registered smell mark in the US (Clarke’s Osewez)²⁴ and over 30 years after the same Sumitomo rose-tyre mark was registered in the UK in 1994 the world’s first smell mark registration.²⁵

Globally, the picture is uneven. The US requires proof of non-functionality and acquired distinctiveness through use, and takes the position that smell marks cannot be inherently distinctive.²⁶ Registered scent marks in the US remain extremely rare. The EU has not registered a smell mark since Sieckmann set a representational standard that available technology could not meet, though the EU’s 2017 revision of the Trade Mark Regulation replaced graphical representation with a broader “clear and precise subject-matter” standard, potentially reopening that door.²⁷ Australia and the UK accept smell marks, broadly on the basis of verbal description for smells that are widely and objectively understood.²⁸

India’s approach combining verbal description with a scientific seven-dimensional graphical representation is the most technologically ambitious attempt yet to provide an objective, reproducible standard for olfactory marks. Whether it succeeds as a durable standard will depend on the refinements that subsequent guidelines and decisions introduce.

Conclusion

The Sumitomo order is an act of principled judicial courage. It reads the statute as it should be read: purposively, with an eye to what trade mark law is for, not merely what its words literally say. It engages seriously with international jurisprudence without being bound by it. It commissions scientific expertise where legal tools fall short. On the central legal question whether the 1999 Act can accommodate smell marks at all? The order answers correctly and with appropriate reasoning.

The criticisms in this article²⁹ are offered in the spirit in which good legal scholarship ought to engage with landmark decisions: not to diminish them, but to ensure that the gaps a first order inevitably leaves do not become entrenched assumptions. Three tests remain. Firstly, the seven-dimensional vector methodology requires refinement through Registry-issued examination guidelines into a publicly accessible, reproducible, and high-resolution standard; this is the Registry’s responsibility. Secondly, the distinctiveness analysis of marks filed on a proposed-to-be-used basis needs doctrinal development through subsequent decisions and, if necessary, legislative clarification; this is the courts’ and the legislatures’ responsibility. Thirdly, the enforcement framework for olfactory rights needs to be built urgently through Rules amendments, practice directions, and the first infringement court’s judgment; this is a shared responsibility across the Registry, the legislature, and, of course, the judiciary.

A registered right without a reliable remedy is, in practice, a paper right. India’s trade mark register has its first smell. Whether that smell can be reliably identified, rigorously protected, and effectively enforced is the question that the decade ahead must answer.