It is a familiar moment in product teams. A device fails in the field, a feature behaves unpredictably, or a process wastes time. Someone isolates the root cause and implements a neat fix that actually works. The next question follows immediately, can this be patented in India even though the underlying product already exists. The short answer is yes, improvement inventions are routinely patentable, provided they clear the twin gates of novelty and inventive step, and do not fall into statutory exclusions. The longer answer is where this article will help.
Improvement patents in India, the legal threshold
Indian law defines an “invention” as a new product or process that involves an inventive step and is capable of industrial application, and “inventive step” as a feature involving technical advance over existing knowledge or having economic significance or both, and that makes the invention not obvious to a person skilled in the art. In other words, your fix must be new and must not be an obvious variation, even if it saves cost or improves performance.
The Supreme Court’s guidance on obviousness
In Bishwanath Prasad Radhey Shyam v. Hindustan Metal Industries (Supreme Court, 1979), the Court emphasised that a patentable advance must be more than a “workshop improvement”. The practical test it applied is whether the improvement would naturally flow to a skilled worker from what the prior art already taught. If so, it is obvious and unpatentable. This remains the touchstone Indian courts and the Patent Office still cite when assessing incremental inventions.
Section 3 filters that catch many “fixes”
Even a new and useful solution can be excluded by Section 3 of the Patents Act. Three filters frequently apply to improvement claims:
Section 3(d) for pharmaceuticals, a new form of a known substance is not patentable unless it shows enhanced therapeutic efficacy. The Supreme Court in Novartis AG v. Union of India (2013) held that improved properties like better stability or flow are insufficient unless they translate into therapeutic efficacy.
Section 3(e), mere admixture, where combining known ingredients only aggregates their properties without a synergistic effect.
Section 3(f), mere arrangement or re-arrangement, where known parts are simply put together, each working independently in the usual way.
For computer-implemented “bug fixes”, Section 3(k) excludes computer programs per se. Examiners look for a demonstrable technical effect or technical contribution beyond the program as such, consistent with the Patent Office’s CRI guidelines.
How examiners and courts evaluate incremental inventiveness
Examiners typically ask, what problem did the art already recognise and what solutions did it point to, would the skilled person have arrived at your claimed path without invention. Delhi High Court decisions in pharmaceutical disputes, including F. Hoffmann-La Roche v. Cipla (2008, 2015), illustrate how Indian courts inspect obviousness arguments, the skilled person’s lens, and the impermissibility of hindsight. While the facts in those cases were drug-specific, the approach is instructive for any incremental invention.
A practical way to frame your improvement is the problem-solution story used globally and reflected in Indian practice manuals, identify the closest prior art, define the objective technical problem, and show how your claimed features achieve a technical effect that the art did not suggest. Doing this in the specification helps the examiner see the non-obvious leap.
Do improvements need to be “better” or just “different”
“Different” is not enough. The claim must be new over all prior disclosures, and the difference must clear the inventive step hurdle. The statutory definition allows “economic significance” to count, however it does not relax the non-obviousness requirement. An improvement that only trims cost using routine substitutions is unlikely to pass.
If someone else owns the base product, can you still patent the fix
Yes. An improvement patent can be granted even when a broader, earlier patent exists. In practice, this may create a blocking or dominant–subservient relationship, where each patentee needs the other’s permission to commercially practise the combined technology. That is a business issue to be solved by cross-licensing rather than a bar to patentability. Please read more about the Freedom to Operate in the market in such cases.
Evidence that strengthens an improvement application
Examiners and courts look for contemporaneous, technical proof that your fix solves a real problem in a non-obvious way. Build the record early by:
Capturing test data that shows a measurable technical effect, for pharma, efficacy data matters because of Section 3(d).
Explaining why prior solutions were inadequate or taught away from your approach.
Avoiding claim language that recites a result to be achieved without specifying the technical means.
Describing the best method you know and enabling others to work the invention, as required by Section 10(4).
Quick map, statute to strategy
Theme | Key law or case | Applicant takeaway |
What makes an “improvement” patentable | Sections 2(1)(j), 2(1)(ja) | Show a technical advance or economic significance and non-obviousness over prior art. |
“Workshop improvement” is not enough | Bishwanath Prasad (SC, 1979) | If the fix flows naturally to the skilled person, it is obvious. |
Pharma improvements | Section 3(d), Novartis (SC, 2013) | New forms need enhanced therapeutic efficacy, not just better properties. |
Mixtures and assemblies | Section 3(e) and 3(f) | Show synergy or integrated interaction, not mere aggregation or rearrangement. |
Disclosure quality | Section 10(4) | Fully and particularly describe the invention and best method. |
Two real-world checkpoints that trip up inventors
1) Prior disclosures that destroy novelty. If you have already put the improved product on YouTube, sold it, or published a paper, novelty may be gone. Limited exceptions exist, for displays at notified exhibitions and papers read to a learned society, but they have strict twelve-month windows. Before any external disclosure, file at least a provisional.
2) The “mere re-arrangement” trap in hardware claims. Many fixes are packaging or layout changes. If each known module keeps working independently in the same way, Section 3(f) is likely to be invoked. The specification must explain the inter-relationship that yields a new technical effect.
How to put your improvement on file, the smart way
File early to secure a priority date, especially if testing or pilots are planned.
Draft around the problem, identify the closest art and articulate the technical effect achieved by your features.
Include comparative data where feasible, and for pharma, link improved properties to efficacy, not just bioavailability.
Consider a divisional if your fix has multiple inventive strands.
Plan for licensing if a base patent exists, a cross-licence may be the commercial solution to a blocking posture.
Can I patent the fix if I do not own the original product?
Yes. Patentability turns on your claims and the prior art, not on who sells the base product. However, practising your improved version might still read on someone else’s valid claims, which calls for a licence, assignment, or design-around strategy. Think “right to exclude” not “right to practise”.
What if I have already disclosed the fix?
Most public disclosures before filing are fatal. Narrow carve-outs exist, for example display at a government-notified exhibition or a paper read before a learned society, provided you file within twelve months. There is also a limited exception for reasonable public working for a trial within one year. When in doubt, treat public disclosure as risky and file first.
Do I need a prototype to file?
No prototype is required. The law demands a complete specification that fully and particularly describes the invention, how it works, and the best method known to you. Detailed drawings, embodiments and test results make your file stronger and speed prosecution.
A final word with case-law perspective
Indian doctrine is friendly to genuine improvements, but exacting about how they advance the art. The Supreme Court’s line in Bishwanath Prasad keeps out routine workshop tweaks. Delhi High Court’s Roche v. Cipla decisions remind applicants that courts scrutinise what the skilled person would glean from prior art without hindsight. For pharmaceuticals, Novartis sets the heightened efficacy bar for new forms under Section 3(d). If your fix clears these gates, is well documented, and is drafted with the problem–solution narrative, an Indian patent is within reach.
