PM Announces a Bold Manifesto
August 15 this year was the 80th Independence Day. There is always a speech at the Red Fort, and I read this one with a mix of attention and skepticism - the speeches tend toward aspirational catalogues that age poorly by the next review. But this one was different in structure, even if the habit of announcement was familiar.
PM Modi named technology as the third “shakti” in his seven-point Viksit Bharat framework. He said AI skill training would be provided to 1 crore - that is 10 million - young Indians within the next year, under the IndiaAI Mission. He said India would add 7 to 8 more semiconductor plants in the next 7 to 8 years, building on the three manufacturing facilities that have already begun production. He called for Made-in-India 6G technology to reach every corner of the country. And he set a figure that was blunt and surprising: 100 GW of nuclear power by 2047.
India of the late 21st century is bound to be a power-hungry beast. Every extractable joule will count. But India’s current nuclear capacity is roughly 7.5 GW. Getting to 100 GW by 2047 means adding more than 4 GW of nuclear annually for the next 20 years, on a technology that typically takes 10 to 15 years per plant to build.
Either the timeline assumes Bharat Small Modular Reactors operating at unprecedented speed, or the ambition is real but the number is aspirational. While I lean toward the latter, the direction of travel is not aspirational - it’s strategic. You don’t get to a semiconductor civilization without reliable, 24/7 baseload power. Solar and wind can’t do that alone. Modi’s chip-AI-6G-nuclear package is not four separate announcements; it is a coherent supply chain argument for what a technology civilization actually requires.
The IndiaAI Mission had made 45,000 GPUs available through its shared computing platform by June. By August, 237 projects had consumed 93.18 lakh GPU hours through that system, and the mission had selected 20 indigenous foundation-model proposals - 12 large multimodal models and 8 small language models - from 506 applications received. These are concrete, real world numbers and we have come a long and happy way since I first read about the at-the-time theoretical IndiaAI mission last year.
India’s 6th Defense Indigenisation List: The One That Covers the Su-30MKI
On August 18, the Department of Defence Production notified the sixth Positive Indigenisation List. PIL-6 covers 405 items: line replaceable units, sub-assemblies, spares, components, and raw materials, with an estimated business potential of 3,070 crore rupees, all of which must be sourced only from Indian industry by December 2031.
A Positive Indigenisation List (PIL) is an Indian government notification that bars the import of specified defence items after fixed deadlines, mandating domestic procurement.
To all the Indian entrepreneurs out there, this list deserves a careful reading. The Su-30MKI is there. The T-90 main battle tank is there. The BMP-II infantry fighting vehicle is there, along with the Advanced Light Helicopter, the Light Utility Helicopter, and the Light Combat Aircraft.
All of these are Soviet/Russian technologies that we have imported for decades, creating a deeply entrenched dependency. The Su-30MKI has been in Indian Air Force service since 2002. Its component ecosystem remained heavily Russian-dependent for more than two decades, a chokepoint that the events of 2022 onward made visibly uncomfortable.
PIL-6 is an exercise in strategic decoupling.
We have now published six of these lists since the Srijan portal went live in August 2020. More than 15,700 defense items have been indigenised since, with an estimated import substitution value of 9,000 crore rupees. It means that roughly one billion dollars of annual defense imports are being redirected into Indian industry, every year, with room for growth.
Back to ISRO
After seven months, ISRO is going back to the launch pad.
The PSLV-C62 mission carrying EOS-N1 failed on January 12, 2026. It was followed by the longest inter-launch gap ISRO has had since the early 2010s - a methodical, unhurried investigation into what went wrong on that morning above Sriharikota.
On August 18, Business Standard reported that GISAT-1A, also designated EOS-05, is being prepared for a September launch aboard the GSLV. It is a high-frequency Earth observation satellite, designed to image large sections of the Indian landmass quickly and repeatedly, with applications that range from flood monitoring to agricultural assessment to forest-change detection.
The pace of the investigation is not exactly comfortable - there are commercial commitments downstream, and India’s launch services ambitions depend on a functioning national launch cadence. But arguably, this is how responsible engineering bodies should operate after an anomaly. ISRO investigated, did not announce a timetable until it was confident, and is now returning.
In July, Skyroot Aerospace’s Vikram-1 became the first Indian private orbital rocket, joining the United States and China as the only countries with domestic private orbital launch capability. The overlap of ISRO’s methodical pause and Skyroot’s commercial speed bodes well for India’s launch ecosystem - it is indicative of a maturing environment because there is now more than one tier of it.
At the National Space Technology Conclave 2026 on August 18, ISRO’s Dr. Nilesh M. Desai, former Director of the Space Applications Centre, laid out a roadmap for what comes next in Indian satellite design:
AI and machine learning for onboard processing and edge computing in orbit,
quantum technology for secure satellite communications covering financial, military, and diplomatic channels,
an expanded radar frequency footprint that now spans almost all major bands.
The immediate implication of that increased radar breadth is that India has become an attractive technical partner for NASA, JPL, CNES, and JAXA on programs that require frequency diversity. We are not just a cheap launch option anymore. We are building systems intelligence.
Solar
India has surpassed Japan to become the world’s third-largest solar energy producer, generating 1,08,494 GWh of solar power against Japan’s 96,459 GWh. That is a 12.5% gap, and it widens every quarter. Only China and the United States generate more solar electricity.
In H1 2026 alone, India added 29 GW of new solar and wind capacity, a 38% increase over H1 2025. The full-year projection has crossed 50 GW of solar additions, which would beat the previous record of 49 GW set in 2025, which was itself a record. Gujarat added 7.6 GW of solar in H1, accounting for 29% of national additions. Rajasthan added 6.6 GW. Under PM Surya Ghar, rooftop solar installations more than doubled year-on-year, adding 6.4 GW in H1 alone.
For context, Germany’s entire installed power capacity - all of it, every source - is about 250 GW. India is adding more than 10% of that figure in new renewable capacity in a single half-year. The electricity generated by India’s solar fleet is not an abstraction. It is factories running shifts in Tamil Nadu, EV chargers in Bengaluru, irrigation pumps in Rajasthan pulling water from depths that monsoons alone could not reach.