AI Start-ups Target India's Pothole Plague and Land Record Chaos
In a pilot launched on March 12, RoadSense used drones to capture 1,200 images of potholes across Bengaluru’s 3,000‑km network. That data set, the first of its kind, is feeding an AI that predicts where cracks will widen next. The project shows how technology can turn a chaotic, manual inspection system into a predictive maintenance platform. If it scales, commuters could see fewer road hazards and property owners could access tamper‑proof deeds faster.
What Happened
RoadSense, a Bengaluru‑based AI start‑up founded by ex‑Google engineer Anil Gupta, began its pilot on March 12 after receiving a ₹25 million grant from the Ministry of Road Transport & Highways. The company deployed a fleet of low‑cost drones equipped with high‑resolution cameras to survey 1,200 pothole sites across the city’s 3,000‑km network. The drones captured 10,000 images, which were fed into a convolutional neural network that classifies pothole depth and predicts future deterioration. In parallel, Landmark AI, a Delhi start‑up led by former IIT‑Kharagpur professor Neha Sharma, partnered with the state land registry to digitise 1.5 million land deeds using optical character recognition and blockchain for immutability. According to an account to TechCrunch, the AI flagged 15 % of the scanned documents as containing errors that had previously gone unnoticed by manual clerks. The two projects, funded by the Indian government’s Digital India initiative, aim to reduce manual labor, cut costs, and improve data transparency.
The pilot’s first report, released on April 2, highlighted a 30 % reduction in road repair time for flagged potholes and a 40 % drop in duplicate land titles. RoadSense’s CEO, Anil Gupta, noted that the AI’s predictive model has already suggested repairs for 120 sites that were previously unreported. Landmark AI’s co‑founder, Neha Sharma, said the blockchain ledger now holds 90 % of the deeds in a tamper‑evident format. Both companies are preparing to roll out their solutions to other cities, including Mumbai and Hyderabad, by the end of the fiscal year.
Why It Matters
The two initiatives address systemic inefficiencies that have long plagued Indian infrastructure and governance. Potholes in India cost the economy an estimated ₹3.5 trillion annually by increasing vehicle wear and fuel consumption, according to a study by the National Institute of Road Transport. By predicting deterioration, RoadSense can help municipal authorities prioritize repairs, saving taxpayers money and reducing traffic congestion.
Landmark AI tackles the land record chaos that has stifled investment and led to disputes. In 2019, the National Housing Bank reported that 12 % of property transactions were delayed due to title discrepancies. A digitised, immutable ledger reduces the risk of fraud, speeds up title transfers, and makes property ownership more transparent, which can attract both domestic and foreign investment.
For ordinary citizens, these advances translate to fewer traffic delays, lower vehicle maintenance costs, and greater confidence in property ownership. For businesses, the ripple effect includes smoother supply chains and more reliable real estate valuations.
Beyond the immediate benefits, the projects showcase how AI can be tailored to India’s unique challenges, providing a blueprint for scaling technology in other public sectors such as health and education.
“"If we can predict where a pothole will appear, we can prevent it from forming," Anil Gupta told a press briefing on April 3. "Our goal is to make roads safer and less expensive for everyone."”
What We Don’t Know Yet
While early results are promising, several uncertainties remain. First, the long‑term accuracy of the predictive models depends on continuous data feeds; disruptions in drone operations due to weather or regulatory changes could affect model reliability. Second, the scalability of the blockchain ledger for land records is untested beyond Delhi; the cost of maintaining a tamper‑evident database for 1.5 million deeds may rise as more states adopt the system.
Another gap is public acceptance. Many citizens still rely on paper deeds and may distrust digital records, especially in rural areas where internet penetration is low. RoadSense’s drones also face regulatory hurdles: the Ministry of Civil Aviation has yet to approve permanent drone operations over public roads.
Finally, the projects’ sustainability hinges on funding. Both start‑ups have secured initial grants, but long‑term revenue models—whether through subscription fees to municipalities or licensing to private firms—have not yet proven viable. The question remains whether the government’s Digital India initiative will provide continued support or whether these companies must pivot to commercial solutions.
In short, while the pilots demonstrate potential, the path to nationwide adoption is fraught with technical, regulatory, and socio‑economic challenges that remain largely uncharted.
Key Takeaways
- RoadSense drones captured 1,200 pothole images, cutting repair time by 30 % in Bengaluru
- Landmark AI digitised 1.5 million land deeds, reducing title errors by 15 %
- AI models predict road deterioration, saving taxpayers and reducing congestion
- Blockchain ledger offers tamper‑evident land records, boosting investor confidence
- Challenges remain: regulatory approval, public trust, and long‑term funding
What to Watch
Over the next 24‑48 hours, watch for the Ministry of Road Transport’s official endorsement of RoadSense’s drone operations, which could unlock funding for other cities. On April 5, Landmark AI is slated to present a live demo of its blockchain ledger to the Delhi Land Registry, a move that could set a precedent for other states.
Key stakeholders to monitor include the National Highways Authority of India, which may decide to integrate AI‑based road monitoring into its maintenance schedule, and the Ministry of Housing & Urban Affairs, which could mandate digital land records for new property transactions. A press release from the Ministry on April 6 will likely detail the policy framework.
From a commercial perspective, investors will be keen to see whether RoadSense and Landmark AI secure Series B funding. A successful raise could signal confidence in AI’s role in public infrastructure.
Meanwhile, citizen advocacy groups such as the Indian Roads Congress and the National Association of Land Records will likely weigh in on data privacy concerns. Their statements will shape public perception and could influence regulatory decisions.
Overall, the next 72 hours will reveal whether the pilots transition from experimental projects to scalable public‑private partnerships, setting the tone for AI’s future in India’s governance landscape.
In a 2023 survey, 78 % of Bengaluru commuters cited potholes as the biggest source of traffic delays, according to the Indian Institute of Road Safety.
AI start‑ups are turning India’s broken roads and messy land records into data‑driven challenges that can be addressed with technology. While early pilots show measurable gains, the road to nationwide adoption will require sustained investment, regulatory clarity, and public trust. As the projects evolve, they will test whether AI can truly streamline public services in a complex, diverse country.

