Supporting a healthier and greener India with our AI

At Google, we’ve long believed in the transformative potential of AI to address some of the most pressing challenges facing humanity. This belief has underpinned our continued efforts and commitment to build AI that benefits everyone.

For the past five years, our research teams have worked to bring this commitment to India, and we’ve been deeply encouraged by the enthusiasm and support with which the ecosystem here has partnered with us – adapting our AI to local contexts to help strengthen communities and improve lives.

We’ve even highlighted how India can maximize AI’s potential in our ‘An AI Opportunity Agenda for India’ whitepaper.

Today, we’re excited to announce new collaborations across health, agriculture and sustainability sectors that will deploy our AI across a range of applications: from helping prevent blindness amongst people with diabetes, to supporting field-level insights that eventually benefit farmers, and even reducing strain on the country’s landfills.

Assisting screening and detection for diabetic retinopathy

Diabetes poses a significant health challenge to millions of people across the world, with one of its most common complications, diabetic retinopathy, resulting in blindness if left untreated. Nearly a decade ago, researchers at Google started exploring the potential for AI to screen for diabetic retinopathy, and to help prevent blindness through early detection, eventually conducting the first patient screening in Madurai, India.

Since then, our diabetic retinopathy AI model has helped support more than 600,000 screenings in clinics around the world. The model supports clinicians with early detection at scale, and in turn assisting timely medical intervention.

And now, Google has licensed this AI model to healthcare providers and health-tech partners Forus Health and AuroLab in India, and Perceptra in Thailand, to support approximately 6 million AI-assisted screenings for diabetic retinopathy in resource-constrained communities in India and Thailand over the next 10 years, at no cost to patients.

We aim for these newest partnerships to contribute to strengthening the global network of innovators working tirelessly to eradicate preventable blindness due to diabetic retinopathy, helping shape a world in which the latest advances in tech enable everyone to see a brighter future.

K.Chandrasekhar, Founder & CEO, Forus Health: “Forus Health is thrilled to partner with Google to bring AI-powered diabetic retinopathy screening to the forefront of eye care by leveraging our innovative retinal cameras and platform. We are confident of impacting millions of lives and fulfil our mission of eliminating preventable blindness.”

Mr R.D. Sriram, Managing Director, AuroLab: “It has been great to see the multiple phases of Google’s diabetic retinopathy AI model - from development to validation to deployment. At AuroLab, we are hopeful that this AI model will play a crucial role in preventing unnecessary vision loss for millions of people.”

Supporting plastic recycling and the circular economy

Plastic recycling is a key lever in climate action, supporting circularity that enables reduced emissions while appropriately and responsibly maximizing use of valuable resources. To assist wider adoption of recycling practices, subsequently contributing to the circular economy, our teams developed CircularNet – an open-source machine-learning computer vision model.

Trained on global datasets and powered by TensorFlow – Google’s free and open-source software library for machine learning and artificial intelligence – CircularNet provides pixel-level instance segmentation, characterizing plastic materials by form and type, to improve how they are sorted, managed, and recycled across the entire waste management cycle.

Saahas Zero Waste (SZW), a Bangalore-based environmental and social enterprise, has been leveraging CircularNet in its efforts to sort plastic waste, improve recycling, reduce strain on landfills from recyclable waste, and strengthen India’s circular economy.

Being used for its material identification capabilities, CircularNet is supporting SZW as a vision-based quality control system that checks quality and quantity of waste before it is sorted, baled and sent to recycling centers, as well as when it is received at the recycling centers.

In a pilot of CircularNet’s AI model at its material recycling facility, Saahas Zero Waste estimates that it realized approximately 85% accuracy in detection of plastic waste. Based on these internal evaluations, SZW predicts the higher quality recyclable materials that are recovered could translate into a 10-12% improvement in revenue generation. They are hopeful that such improved financial viability for material recovery facilities could stand to benefit the many other entrepreneurs in India who operate in this space. Saahas Zero Waste has further estimated that roughly 90% of recyclable waste could be diverted from landfills.

Arun Murugesh, Vice President, Sales and Marketing, Saahas Zero Waste: “Saahas Zero Waste is committed to the evolving waste management ecosystem in India, aiming to responsibly manage over 500 tonnes of waste daily by 2026. Google’s CircularNet model is playing an important role by supporting automated AI-powered quality control systems that can be deployed efficiently at scale - and we’ve been greatly encouraged by early results. Improved quality assurance at waste recovery facilities and recyclers powered by CircularNet promises to improve resource recovery through closed loop recycling. By combining technology with viable business models, we aim to contribute towards India’s vision of a circular economy. We also see AI-assisted partnerships as a driver to achieving larger impact, including livelihood generation, safer working conditions and greater ecological conscientiousness.”

Strengthening the agriculture ecosystem with field-level insights

Insights in the agricultural ecosystem today exist at an aggregate level, with current solutions working effectively at village, sub-district or block levels. But, if they are to benefit the millions of India’s farmers facing challenges at myriad levels - from low yields to the rising threats of climate change - agricultural solutions need to be cost-effective, simple, and especially, targeted.

This is what the Agricultural Landscape Understanding (ALU) Research API, announced at I/O Connect in Bangalore earlier this year, was designed to help provide. ALU API uses remote sensing and AI to provide insights at individual farm level across India. By combining high-resolution satellite imagery with advanced machine learning models, the ALU API can identify fields, water bodies and vegetation boundaries and their acreage, organizing agricultural information and providing it as a base layer for the agri ecosystem.

We are now opening up access to the ALU API for developers. This will enable support to stakeholders across the agri ecosystem in their respective efforts: from helping public sector agencies improve the accuracy of land records, to agri-input companies optimizing the distribution of resources like seeds and fertilizers. Agritech startups may also consider using it to develop precision agriculture tools that provide tailored advice to farmers based on their unique field conditions.

Such capabilities can assist more data-driven and efficient decisions across India’s agri ecosystem, while supporting the development of precision agriculture tools, optimizing resource allocation, and improving agricultural practices that help farmers improve farm management and productivity.

The partnerships announced today build on Google’s continued commitment to collaborative innovation and to seeing the advantages of AI translate into positive impact across India. We look forward to deeper partnerships across sectors that meaningfully contribute to a more sustainable and inclusive future for everyone.