We have to make industrial AI robust: Dr. Peter Koerte, Siemens

Transform -- Siemens Innovation Day 2026 was held in Mumbai today. Sunil Mathur, MD and CEO, Siemens Ltd welcomed the audience.

He said six years ago, we weren't sure how this would develop. Government has laid out some points for the country to move forward. We are currently seeing trade realignment, weaponization of the critical resources, eg., rare earth minerals, supply chains, etc., and an AI explosion.

AI has become much more real in the industrial manufacturing space, and in day-to-day lives. India has low compute cost (India AI MIssion), rail electrification, with largest green rail network if 69.4 km, mobile data consumption, with fastest 5G adopter, and ~85% population network covered, and clean energy, with over 50% power from non-fossil fuels, and five years ahead of COP26 target. The country has been delivering on all of these programs.

India is now upping the game. Innovation will have to play central role in growth story. We have to move up from 0.3% to 3% in R&D. In data centers, we will need to increase from 1.3% to 8%/. Renewable energy should rise from 250GW to 500GW. Green hydrogen production should rise from 0.25 to 5 MMTPA. Industrial corridors should move from four in 2025 to 11 by 2030.  Manufacturing growth should move from 16-17% in 2025 to 25% by 2047. 

There is lot of potential in all of these areas. The world is currently volatile. You want to scale, in sustainable manner, using AI, and technologies that are secure. There are technologies that will create the future. These are: industrial AI, digital twins and digital threads, autonomous buildings, digital substation, next-gen automation, next-gen grid management, signaling in cloud, etc.

Dr. Peter Koerte, Member of the Managing Board, CTO, and Chief Strategy Officer, Siemens AG, presented the keynote. He first congratulated everyone on India's win over England yesterday at the T20 World Cup Cricket.

He talked about industrial AI, and how it changes the world. We are talking about grids, semiconductors, life sciences, etc. India is growing GDP at 7%. The future is looking absolutely bright for India. It is going to gain and contribute the most. We will talk about AI revolution in industry, India's unique position, and from potential to impact.

AI and Nvidia, and chips and GPUs, are always part of conversations. There are 20 billion smart devices out there. In factories, you can get TB of data in an hour. Data can be used to feed the AI. About 80% of data today is currently not used. We are focused on making the industry much more intelligent and effective. We are combining the real with the digital. You can simulate what is happening. Automotive is a key area for innovation, as is aerospace.

At Audi, we have cameras to detect and identify welding spots that are accurate. It saves lot of money, and ensures quality. Software-defined automation is also gathering steam. For infrastructure, water is key for India. You get 60% of losses as it is transported to the users. We are cutting that down. For buildings, 30% of all energy consumed today is used to make us feel comfortable. We are cutting down on the energy cost gradually, and make billings more efficient.

Robust industrial AI
That's brings us to industrial AI. It needs to be safe, secure, reliable, and trustworthy. The answer could be currently 80% correct. We need that to be 95-99% correct. We have to make industrial AI robust. It can be deployed across all different areas. There are tariffs coming up all the time, trade conflicts, etc. We need to adapt to the dynamic, ever-changing environment. 

India is very strong in software development. It can become a major force in AI. It can also apply AI across the various industries. Around 87% of Indian enterprises are already using AI solutions. About $152 billion is to be invested in semiconductors and manufacturing. We need to also leapfrog, and make all that AI native. Lot of apps are also going to come from India for the world.

AI, made in India, is now exported globally. It is adapting the operations to extreme weather. Seven+ days are won for planning ahead, along with +70% accuracy in predicting land accumulations, and +20% cost savings for customers.

Siemens Xcelerator has curated portfolio, has powerful ecosystem, and has growing marketplace. All hardware and software can be found here. For example, the oil and gas customers can definitely find success, especially during sandstorms. 

Siemens now wants to play a role, and lead in industrial AI. It is no. 1 in industrial automation, grid protection and automation (IEC), rail infrastructure, etc. We have 1500+ AI experts, and over 50 years of applied AI experience within Siemens, We are no. 1 in industrial software. 

Industrial AI now needs data! We can build industrial AI by doing it all together. We may have 300 million CAD files of data. But, it may not be enough to build a real-world product. We are serving over 30 different industries today. For data, nobody can do it alone, and, nor can it do for technology. We are also working with Nvidia. We are helping customers build the digital world much faster. You can build ships having over 7 million parts. Simulations can allow much faster time-to-market.

Industrial AI for India
Vishal Dhupar, India head, Nvidia, said there's no second thinking on our part, as manufacturing has to be done much more. Nvidia is helping Siemens scale further. Number of greenfield projects are coming up and increasing. PLIs have already generated 1.2 million jobs. Third, we have the AI flywheel. Government is building AI brains here. Combining all of these three brings industrial AI, for India.

Sangeet Kumar, CEO and Co-founder Addverb, stated that it is a robotics company, and builds all forms of robots. We cater to industrial and defense. We need Nvidia and Siemens as partners for industrial AI. With Siemens, we started in 2018. Everything happens on the Siemens TeamCenter platform. Simulation is done on PlantSim. We sell solutions to customers that can solve their problems, using Technomatix. We also use the Mendix platform to help customers.

As for Nvidia, we have cobots, etc. All those training of the models are done on the Nvidia simulation platform. We are building great, reliable products for our customers with the help of Siemens and Nvidia.

Dhupar added we are now building industrial AI OS. You have optimization taking place, and secure supply chains. Reliance New Energy is virtually building an AI factory. They are accelerating the entire design cycle, based on Siemens and Nvidia. We are also going to build 12 brand new industrial cities. Energy needs to be added to that. We need to build real-time scenarios. In Pune, there is a lab that can do crash analysis. We are looking at AI moving to the world of atoms, and being grounded there.

Kumar noted that India is very important for Addverb. We initially wanted to help the Indian companies become more productive. Today 40% of our revenue comes from India. We are scaling up now. We export robots from India to over 25 countries today. In the next decade, we need to reach out to over 100 countries from India.

There is also an opportunity in green energy. Reliance New Energy has 20GW factory. There are several solar factories, with at least 11 of those as our customers. Apple's OEMs such as Tata and Foxconn are also our customers. Today, every country would like to have lots of industries within their own country. 

Dr. Koerte noted there are so many proof points that do not scale. We have to do this first, customer by design, and then build applications. Use case may scale, else, it may remain a PoC. 

He added at a presscon that Siemens has around 4,700 IPs. AI is going to be great support for SMEs. The SMEs are not yet ready. We are building components for the industry. We do not currently manufacture robots. We develop the programming software.