Indian Physical AI attracts $155 Mn in 2026 as robotics, real-world data gain traction

Home services startup Pronto’s decision to expand its Physical AI experiment involving wearable cameras inside customers’ homes has brought the emerging technology into focus, particularly the opportunity around real-world data for training AI and robotics systems. The move comes even as larger home services players Snabbit and Urban Company's InstaHelp have opted out of similar initiatives. Pronto, which has raised $45 million in funding this year alone, is expanding its opt-in Pronto Verified service, which records service sessions and uses anonymized footage for AI training. As companies explore ways to collect real-world data for AI and robotics, investor interest is also growing across robotics, autonomous systems, AI-powered hardware, and training infrastructure. Year-on-year funding trend According to data compiled by Entrackr, startups in the Indian Physical AI ecosystem have raised around $155 million across 31 deals in 2026 so far. This compares with $130 million across 25 deals in 2025, $124 million across 24 deals in 2024, and $91 million across 21 deals in 2023, according to Venture Intelligence data . For this analysis, the funding dataset includes startups headquartered in India, as well as overseas companies with significant Physical AI product development, commercial deployment, or data collection in India. Companies with only engineering or support teams in India, or where Physical AI remains an experimental or ancillary initiative, such as Pronto, are excluded from the funding tally. Funding in 2026 so far is already about 19% higher than the total raised in 2025 and more than 70% above 2023 levels. Most companies that raised capital this year are early-stage ventures, though capital remains concentrated among a few larger rounds. Innefu Labs, Unbox Robotics, and Deccan AI together raised $83 million, accounting for more than half of the $155 million raised so far. Perhaps predictably, this remains the safest bet on the AI boom, in terms of predictable, if relatively tiny part of it. With firms building their own layers on top of existing frontier AI networks or otherwise. Some would say this is just as well, pointing to the havoc the rise in valuations, followed by the crash it has caused in South Korea this month. Growth-stage deals lead the pack Innefu Labs led the segment with a $30 million Series B round led by Panthera Growth Partners. The national security-focused AI company is also building a dedicated Physical AI and robotics division for defence and security applications. Supply chain robotics startup Unbox Robotics raised $28 million in a round led by ICICI Venture, with participation from RedStartLabs, F Prime, 3one4 Capital, Navam Capital, Force Ventures, and existing investors. Deccan AI raised $25 million in a Series A round led by A91 Partners, with participation from SIG and Prosus Ventures. The startup provides data and model training, post-training data, and evaluation services for frontier AI labs, and has expanded into enterprise AI solutions. Outside the funding dataset, US-based physical security AI startup Hakimo , founded by Indian-origin entrepreneurs Sam Joseph and Sagar Honnungar, raised $12 million in a growth round led by Zigg Capital, with participation from Neotribe Ventures, Vertex Ventures, Defy.vc, and Rocketship.vc. It uses AI to automate physical security monitoring through existing camera infrastructure. Early-stage startups gain momentum Marine robotics company Rekise Marine raised $9.7 million in a seed round co-led by Accel and NKSquared. It builds autonomous ships and submarines and partners with Indian shipyards for vessel design and integration. San Francisco and Bengaluru-based Human Archive raised $8.2 million in a round led by Wing Venture Capital and NVP Capital. The Y Combinator-backed company came into the spotlight in India following early talks with Snabbit around a potential Physical AI initiative, though the home services startup eventually decided not to proceed . Human Archive is building large-scale human sensorimotor datasets to train robots and Physical AI systems. Bengaluru-based SwitchOn raised $8 million in a pre-Series B round led by IvyCap Ventures. The startup develops AI-powered vision systems for automated quality inspection across manufacturing production lines. Other notable rounds include Spector.ai’s $6.7 million raise led by IvyCap Ventures and ANSCER Robotics’ $5.4 million (Rs 45 crore) Series A led by IAN Alpha Fund, with participation from Info Edge. Bengaluru and Detroit-based Mowito raised $3 million in a pre-seed round led by Version One Ventures to build AI foundation models that enable industrial robotic arms to learn manufacturing tasks through demonstrations. Construction-focused Physical AI and robotics startup Flo Mobility raised $2.5 million in a pre-Series A round co-led by Mela Ventures and Arali Ventures. The remaining funding was spread across more than 20 deeptech and robotics startups, including AutoVRse, WorkOnGrid, RoshAI, Armatrix, Constems, and others. The investor base across these deals includes mainstream venture and growth funds such as Accel, A91 Partners, ICICI Venture, IvyCap Ventures, 3one4 Capital, Prosus Ventures, SIG, Info Edge, and Version One Ventures. The data layer around Physical AI A separate set of companies is working on the data collection, annotation, and training infrastructure required for Physical AI systems. Human Archive is building human sensorimotor datasets for robots and Physical AI systems, while Neocambrian AI has launched an India-focused robotics data factory. Other companies working across this layer in India include iMerit, Human Stryde, Luel, Awign, Build, Centific, Humyn Labs, Aura ML, Build AI, FPV Labs, and Northstar. Physical AI models require real-world data that captures human actions, objects, environments, and interactions. Much of this data has to be collected specifically for robotics and embodied AI use cases rather than sourced from existing internet datasets. Pronto offers another example of how such data could be generated. While the company is not included in the $155 million Physical AI funding tally, its experiment shows how businesses with workers operating in real-world environments could become a source of training data. Larger rounds could follow Skylark Labs, led by India-born founder and CEO Amarjot Singh, has a significant India presence and works with the country’s defence ecosystem, including the Indian Army, Indian Air Force, and ideaForge. The company is in talks to raise $100–150 million at a valuation of more than $1 billion. If completed, the round would be substantially larger than any individual Physical AI funding round involving an India-linked company in 2026. Physical AI companies, particularly those building hardware, will still have to deal with higher deployment costs and longer commercial cycles. For companies collecting real-world data, privacy and consent will remain key concerns, especially for data collected inside homes.