Artificial intelligence (AI) can be a game changer for India’s rapidly expanding distributed renewable energy, said JVN Subramanyam, Joint Secretary at Ministry of New and Renewable Energy (MNRE) at India AI Impact Summit at Bharat Mandapam on Monday.
His comments came during a discussion titled ‘Global Mission on AI for Energy Scaling through citizen-centric India Energy Stack’.
The event saw participation from Hemang Jani, who represents India, Bhutan, Bangladesh and Sri Lanka at the World Bank, and Henri Verdier, director general of the INRIA Foundation in France.
Distributed renewable energy (DRE) refers to small-scale, decentralised power generation systems — typically ranging from a few kilowatts to megawatts — that produce electricity from renewable sources directly at or near the point of use, such as rooftop solar, small wind turbines, or biomass.
Subramanyam said 52% of India’s total installed power capacity is currently sourced from non-fossil fuel sources. That is around 272 gigawatts.
Out of this, at least 140 gigawatts of India’s installed capacity is solar. Within this, 38 gigawatts comprises DRE.
Over the last 15 months, India has added almost 18 gigawatts to the distributed renewable energy space both under the Pradhan Mantri Surya Ghar Muft Bijli Yojana, the rooftop solarisation programme, and the Pradhan Mantri Kusum Yojana, he said.
Story continues below this ad
“We spent close to 9 billion USD on rooftop solarisation and about 4 billion dollars on PM Kusum Yojana,” he said.
He said this scale-up in the past year was enabled by technology solutions that have been incorporated and which benefit consumers, field workers, vendors, banks and DISCOMs alike.
Looking ahead, Subramanyam said AI will play a transformative role in managing the next stage of distributed renewable energy growth.
Pointing to structural constraints in the distribution network, Subramanyam noted that transformers were originally designed for unidirectional power flow, but now must accommodate millions of ‘prosumers’ — consumers who also generate electricity. As distributed renewable energy expands, demand response management and asset maintenance across geographies will become increasingly important.
Story continues below this ad
“Can we build predictability into this? There are a lot of generators who require weather forecasting, predictive analytics. Can we get this onto the AI tool? Lastly, as a consumer, as a prosumer, whether I know it is my rooftop system, is my solar pump generating enough when compared to my peers? Can I be able to sell my electricity to the other peers, which the B2B is trying to enable? So, these are enabling decisions,” he said.
“So, if you ask me what the AI is going to do in the future, it will not just make us react, but it will make us act. So, that is the story that we are looking forward as far as DRE and AI convergence is concerned,” he added.
AI as Development Infrastructure
Hemang Jani said AI must be treated not merely as a technological tool but as development infrastructure itself.
“The way we look at the grid, the way we look at discoms, and the way we look at meters, we have to look at AI. That is also one of the infrastructure layer that we need to create. And that is where I think the game will change,” he said.
Story continues below this ad
He also emphasised that countries cannot afford to treat AI as a pilot exercise but must deploy it at scale, citing the India Energy Stack as an example of thinking big. He said India should aim to become “the Google of AI for energy, for the world,” positioning itself as a global leader in AI-driven energy systems.
Henri Verdier underlined that the accelerating energy transition — marked by rising renewable penetration, decentralisation and growing consumer expectations — is making power systems far more complex, and artificial intelligence may be essential to managing that shift.
However, he cautioned that innovation does not automatically mean progress, pointing to past digital missteps such as poorly regulated social networks that enabled Big Tech dominance.
Open standards like TCP/IP spark wide innovation and teamwork in diverse ecosystems, while strong rules and open-source AI prevent a few giants from ruling via subsea cables or models — favoring local solutions for energy, farms, and more.
Story continues below this ad
“We cannot afford a world with a few AI giants controlling the effects of the economy,” he said.
Defining Success
On what would constitute success in the near term for AI in energy, Jani said it is a fast-moving and evolving technology, and policymakers should not look for short-term wins but long-term transformation.
Meanwhile, Subramanyam said in the next two to three years time, he sees the success in AI and RE convergence would mean “where overall cost of power to the consumer goes down and our industrial competitiveness goes up and consumer empowerment becomes prosumer empowerment and our grids become ready for energy transition.”
.