WEF's Artificial Intelligence for Agriculture Innovation (AI4AI) Project

At the Centre for the Fourth Industrial Revolution India, the World Economic Forum launched the Artificial Intelligence for Agriculture Innovation (AI4AI) Project along with the State Government of Telangana on 12 August 2020. It is the first project of Telangana’s Centre for Responsible Deployment of Emerging Technologies which is a virtual interdisciplinary hub to collaborate on Fourth Industrial Revolution technologies and their pilot deployments within an evidence-based and consensus driven governance framework.

Capitalizing on the value of emerging technologies such as artificial intelligence (AI), blockchain, drones and internet of things (IoT) has the potential to impact productivity and efficiency at all stages of the agricultural value chain – to meet the aspirational goals of doubling farmers’ income and increasing farm productivity while reducing wastage and enhancing supply chain efficiency and transparency.

The project community brings together more than 100 leaders from the agriculture value chain: state and central governments, Indian and global companies, technology innovators, and academic and philanthropic communities with the objective of strengthening multi-stakeholder collaborations to analyse and work on the opportunities and challenges of applying upcoming technologies to transform the agricultural landscape to be profitable and sustainable for farmers.

Agriculture plays a pivotal role in India’s economy as over 58% of rural households depend on it as the principal means of livelihood. The launch of this initiative is a commitment to improve the state of the farmers’ world with the operating principle to “think big, start small and scale fast”. The integration of emerging technology in this domain is the only way we can exponentially grow – from better irrigation and soil health to avoiding pest infestation and waste management in the supply chain, everything can be made possible to the remotest parts of the country. Data is the oil for any new economy – by realizing its criticality and solving existing data ownership and privacy concerns, there is potential to not only impact Indian farmers, but those all over the globe as well. Electronic farm records provide real time data on the health of the farm, covering various aspects related to the farm and farming activities; together with AI and IoT, it will be the key to precision agriculture.

In India, the Ministry of Agriculture has collected land records, soil health and other farm details from five crore farmers to set up a digital agriculture stack. Realizing the importance of AI in agriculture, it extended the National e-Governance plan for Agriculture to fund projects using emerging technologies such as blockchain, AI and machine learning, IoT and others – it is the combinatorial power of these technologies that can aid in solving the problems on the ground.

The Ministry of Electronics and Information Technology has also launched several programmes in this sector, such as soil health cards, crop insurance and loans, and supply of fertilizers and pesticides. The Government of Telangana defined its vision to be a global leader in emerging technology and accelerate AI-led innovation for social impact, especially in the public sector. Under this project, the team has partnered with start-ups to design and deploy solutions, which are being implemented in five districts of the state; hence providing a rich test-bed to learn and accordingly develop a plethora of use cases to then carry out across the country.

Technology interventions have the potential to solve key challenges of our farmers – low farm productivity, information asymmetry, transparency required in supply chain loss etc. The panel divided the use of technologies in agriculture into two parts: on-field interventions (precision agriculture – weather forecasting, autonomous tractors and pesticide delivery by drone) and off-field advances (accurate forecasts of demand and supply, improvement in price discovery and reduction in waste) and the different technology interventions being employed.

The aim should be to provide farmers with a one-stop solution by orchestrating all these services with governments, technology and start-up companies, research and agriculture universities, as well as civil society organizations. The creation of such an alliance will ultimately deliver on the promise of bringing emerging technologies to the farmers and digitalize the agricultural ecosystem.

To deep dive into different aspects of the agriculture value chain, four working multi-stakeholder groups have been set within the alliance structure with the objective to have a more intensive and focussed discussion on the challenges and opportunities discussed at specific stages in the value chain, using a combination of interventions relating to technology governance and deployment of emerging technologies. Each of the groups meets on a weekly basis to further the agenda. Within the next three months, the groups would identify the major issues and opportunities within each area, define the outcomes envisaged, delineate the criteria for assessment of the impact, and recommend policy and institutional arrangements required to ensure the sustainability of the identified interventions. The four groups are as follows (along with broad ranging themes within each):

Intelligent crop planning and sowing

· Enable the farmers to know, in advance, when sowing is supposed to happen and at what time they would be able to gain the maximum market price for their produce

· Start-ups and governments are working in isolation trying to build databases that include farmers’ and farm-related information. Instead, they should combine forces to create an open forum/database with this information and update it on a regular basis, which will help stakeholders to take well-informed decisions; this platform can also include information on soil testing, physical access to seed and other input dealers, as well as market access

· Weather data is available today at a macro level. There is a need for accurate rain predictions to enable farmers to decide when and how much to irrigate and ensure productivity. Incorporating this aspect into precision agriculture, AI-based tools have been developed to compute agronomic insight, such as a crop health stress alert and pest or disease alert.

· Establish a linkage between market visibility in terms of pricing and availability of products and production, so farmers can backtrack and ascertain what to produce

Smart crop health management, precision farming and yield analytics

· The fortune of farmers is heavily dependent on the equipment they use, but also on extensive use of accurate data and sensing. Technology companies and start-ups have employed drones for farm imagery, along with analytics and deep learning, to monitor crops at different stages (from irrigation, soil treatment, fertilizer use and harvesting) without ever visiting the farm. Hence, it is important not only to develop the technology, but also to investigate applying cognitive computing to create precision farming.

· We need to be farmer centric, have ground level understanding and develop comprehensive solutions instead by looking at the chain as an integrated process. Different start-ups, technology organizations and policy makers need to come together to work on bringing about some change in this landscape

· There is a crucial need to build a proper framework and architecture in place around these various processes so they can be scaled up when the time is right

Farm gate to fork:

· Digitally transforming the value chain There are two ordinances in the Essential Commodities Management Act that enable contract farming –purchasing of farm produce at or before the time of sowing, and second, selling at a trade area that is outside the traditional Mandi spaces; a digital extension service is needed so farmers can get better advisories on specific farms, their skill gaps can be mitigated, and they can get information on how to leverage these ordinances

· Digital platform and marketplaces need to be built at state and village levels to enable services to manage services such as logistics and warehousing; governments will need to come together to either build a single platform or lay down proper protocols and specifications for such platforms

· There is a need to have a set of common interoperability protocols evolving through the digital agriculture stack that is ongoing for various services (similar to KYC for opening bank accounts or NPC for payments); such an implementation can then bring on board marketplace providers from the private sector and enable the ecosystem to bring about a big change in the system.

Data-driven agriculture:

· Companies are harnessing the power of IoT, the cloud and AI together to build powerful platforms that reduce time and increase accuracy in both on-field and off-field activities. These platforms aggregate agriculture datasets from different sources, fuse them together and build AI and machine learning models to build custom digital agriculture solutions. There have also been explorations into creating centralized knowledge hubs by augmenting comprehensive datasets with real-time data from IoT sensors, non IoT databases, satellite UAB images to relay intelligence on soil content, pest infestations, moisture content and weather alerts.

· Immediate need is to build a national library of datasets that are openly available, e.g. an open data registry platform; this should include a unified digital land record system for the entire country; a common data model would incorporate all of the data currently lying in isolation into a standardized registry

· Public-private partnership is key to building data as a service, which can be leveraged by ag-tech companies or start-ups to build point solutions on top of this standard common dataset, and making it available to the whole country

Enable strong data governance policies on encumbrance management, auditability and screening to ensure that the right data is used for the right purpose across the entire value chain

Point of contacts-

• Purushottam Kaushik, Head of the Centre for the Fourth Industrial Revolution India

• Shubhangi Poddar, Community Lead of the Centre for the Fourth Industrial Revolution India