From GaN Circuits to CAR-T: India's Deep-Tech Pipeline Deepens While Patents in Force Lag

The news

A commentary in The Hindu by Mukundan Chakrapani and Gaurav Jain argues that three foundations of an innovation economy — public research, corporate R&D and deep-tech entrepreneurship — are now advancing together. In March 2023, DRDO scientists at the Solid State Physics Laboratory, Delhi and the Gallium Arsenide Enabling Technology Centre, Hyderabad announced a breakthrough in gallium nitride monolithic microwave integrated circuits, the know-how for which was, by widely reported accounts, refused to India under the Rafale offset provisions. DRDO is now transferring GaN High Electron Mobility Transistor-based MMIC technology for 5G and 6G infrastructure, EV on-board chargers and renewable energy inverters; India is one of seven countries to have mastered it, with China, France, Germany, Russia, South Korea and the United States. AGNIT Semiconductors, a spin-off from the Centre for Nano Science and Engineering at IISc Bengaluru, is commercialising it. The Bharat 6G Alliance aims to contribute 10% of global 6G patents by 2030; per a review meeting by Communications Minister Jyotiraditya Scindia its members have made more than 7,700 patent filings across 5G and 6G, including over 4,400 foreign filings — applications, the authors stress, not grants or declared standard-essential patents. Almost 3,000 technical contributions were made to 3GPP last year, a fifteen-fold rise from 2020, and WIPO's 2025 Patent Cooperation Treaty rankings placed Jio Platforms 19th, its first top-20 entry. Indian patent filings grew from just over 1,10,000 in 2024-25 to more than 1,43,000 in 2025-26, a 30.2% increase, with domestic applicants at almost seven in ten; but patents in force, just over 2,40,000 in 2025, are a fraction of China's 5.7 million, the United States' 3.5 million and Japan's 2.1 million. The authors cite Pixxel Space, founded by two BITS Pilani alumni, with six hyperspectral imaging satellites in orbit and 18 to 24 planned; Skyroot Aerospace's Vikram-1 launch and Agnikul Cosmos's reusable launch vehicle work; and ImmunoACT, incubated at IIT Bombay with the Tata Memorial Centre, whose NexCAR19 is India's first indigenous CAR-T cell therapy, at a tenth of typical costs. Separately, The Hindu reports that India Deep Tech Alliance members invested ₹2,170 crore in 56 deep-tech firms over the last year, announced at Semicon India 2026; Principal Scientific Adviser Ajay Kumar Sood said this showed the vision behind the Department of Science and Technology's ₹1 lakh crore Research, Development and Innovation scheme translating into measurable action. Energy and climate start-ups took over 30% of the capital, eight firms receiving ₹655 crore, and AI nearly 30%, with 16 firms receiving ₹639 crore.

Static syllabus linkage

1. A filing, a grant and a patent in force are three different things. A patent application is a claim; a grant is the examined and allowed monopoly; a patent in force is a granted patent on which renewal fees continue to be paid, which is the only one of the three that signals commercial value. Under the Patents Act, 1970 the term is twenty years from the date of filing, and applications are published eighteen months from priority. The gap in the op-ed's own figures — filings up 30.2% while patents in force stand at just over 2,40,000 — is the difference between filing behaviour and commercial value, and it is the statistic examiners reward candidates for understanding.
2. The Patent Cooperation Treaty is a filing route, not a world patent. The PCT, administered by the World Intellectual Property Organization, lets an applicant file one international application that preserves the right to seek protection in member states, with national phase entry later. There is no such thing as a global patent; grant remains national. India's patent administration sits with the Controller General of Patents, Designs and Trade Marks under the Department for Promotion of Industry and Internal Trade. WIPO's annual PCT rankings, in which Jio Platforms placed 19th, therefore measure international filing activity, not granted rights.
3. Standard-essential patents are where telecom money is made. 3GPP is the body that develops mobile telecommunications standards, and a patent whose use is unavoidable in implementing a standard becomes a standard-essential patent, licensed on fair, reasonable and non-discriminatory terms. A country that only implements standards pays royalties; a country whose firms hold SEPs receives them. This is precisely why the op-ed distinguishes 7,700 filings from declared SEPs, and why the Bharat 6G Alliance's target is framed as a share of global 6G patents.
4. Gallium nitride is a wide-bandgap semiconductor, and that is the whole point. GaN has a much wider bandgap than silicon, so devices built on it tolerate higher voltages, higher frequencies and higher temperatures. That makes GaN High Electron Mobility Transistors the material of choice for active electronically scanned array radar, satellite communications, electric vehicle chargers and solar inverters. A monolithic microwave integrated circuit fabricates the amplifiers, mixers and switches of a microwave system on a single chip. Because of the defence applications, the technology is tightly export-controlled, which is why indigenous mastery is a strategic and not merely a commercial fact.
5. CAR-T therapy, in one paragraph. Chimeric antigen receptor T-cell therapy takes a patient's own T cells, genetically engineers them to express a receptor targeting a tumour antigen, expands them and reinfuses them. NexCAR19, developed by ImmunoACT with IIT Bombay and the Tata Memorial Centre, targets the CD19 antigen found on certain B-cell cancers. It is a living drug, manufactured per patient, which is why cost has been the barrier globally and why a tenth-of-cost Indian version matters for access rather than for novelty alone.

Why UPSC loves this

1. This is the GS3 indigenisation question in its purest form. 'Achievements of Indians in science and technology; indigenising technology and developing new technology' is verbatim syllabus. The GaN story supplies the causal sequence the examiner wants — denial, public laboratory response, academic talent pipeline, start-up commercialisation — rather than a list of achievements.
2. Intellectual property is examined as an economics question, not a law question. UPSC has asked about India's innovation ecosystem, about why research spending has not translated into commercial outcomes, and about the gap between patents filed and value captured. The filings-versus-patents-in-force contrast, and the SEP point, let a candidate make that argument with numbers rather than adjectives.
3. Space and biotech start-ups are now standard illustration material. Hyperspectral imaging satellites, private launch vehicles and indigenous cell therapy are examples that serve answers on private participation in space, on affordable healthcare innovation, and on start-up policy simultaneously. Carrying four or five such named examples with one specific fact each is more useful than carrying a scheme list.

Analysis

1. Denial is the most reliable innovation policy India has ever had. The GaN MMIC capability was built after the technology was withheld under the Rafale offset provisions; the cryogenic engine, the supercomputer and the nuclear fuel cycle follow the same pattern. This is worth stating precisely because it cuts against the assumption that access to foreign technology accelerates capability. Access often substitutes for capability, because importing is cheaper than inventing until it is not available at all. The policy implication is uncomfortable but real: capability in strategic sectors tends to be built when procurement stops being an option, which argues for treating some import dependencies as deliberate policy failures rather than efficient sourcing.
2. The filing numbers are impressive and the ownership numbers are not. Filings rose 30.2% in a year to more than 1,43,000, with domestic applicants at nearly seven in ten, and Jio entered the PCT top 20. Yet patents in force stand at just over 2,40,000 against China's 5.7 million. A patent in force is one whose owner is paying renewal fees because it is worth holding, so the ratio of filings to patents in force measures whether inventions are reaching markets. India's filing surge is real but recent, and much of it will never be renewed. The authors' own caution — that 7,700 filings are applications, not grants or declared SEPs — is the correct way to read every such number.
3. Examiner capacity is the bottleneck nobody campaigns about. The op-ed names expanding examiner capacity at the patent office as a requirement, and it deserves more weight than it gets. A filing surge without a matching examination capacity converts into pendency, and pendency destroys the commercial value of a patent because the licensing negotiation cannot begin until rights are certain. This is an unglamorous administrative reform of exactly the kind that determines whether the innovation rhetoric produces revenue, and it costs a rounding error compared with the ₹1 lakh crore RDI scheme.
4. Standards participation may matter more than patent counts. Nearly 3,000 technical contributions to 3GPP last year, a fifteen-fold increase from 2020, is arguably the strongest signal in the entire piece. Standards bodies are where the technical architecture of a network generation is decided, and firms in the room shape specifications towards technologies they own. A country that shows up with contributions is bidding to own; a country that shows up with a patent count is bidding to be counted. The 6G Alliance's target of 10% of global 6G patents by 2030 will be met or missed largely in those committee rooms.
5. Public capital and private capital are being conflated in the announcements. The India Deep Tech Alliance's ₹2,170 crore across 56 firms was, on its own statement, invested independently by individual members according to their own strategies — that is ordinary venture capital, not a pooled fund and not a government programme. Presenting it alongside the ₹1 lakh crore RDI scheme as evidence that the scheme's vision is translating into action is a rhetorical move, not a causal claim. The honest reading is that private deep-tech capital has arrived and the public scheme has not yet deployed at scale; whether the second caused the first is unestablished.
6. The allocation pattern tells you what investors think India can win. Energy and climate start-ups took over 30% of the capital across just eight firms at ₹655 crore, AI nearly 30% across 16 firms at ₹639 crore, and quantum, robotics and space split a similar share across 21 companies. The concentration in energy — fewer firms, larger cheques — suggests capital-intensive hardware bets, while the wider spread in quantum, robotics and space suggests earlier-stage exploration. That pattern is consistent with a market that believes India can manufacture energy technology at scale before it can build a quantum computer, which is a sober judgement worth taking seriously rather than resisting.

Administrator's brainstorm