Applied Analytics for Business – Agriculture PDF

Applied Analytics for Business – Agriculture Dr. Ravi Shankar Prof – Data Science (AI -ML) Dean – Enterprise Solutions & Academic Collaborations https://www.linkedin.com/in/ravi-shankar-70584b1/ [email protected] First Session , 29th AUG, 2023 Introduction – Global Agri : Challenges & Solutions...

Applied Analytics for Business – Agriculture Dr. Ravi Shankar Prof – Data Science (AI -ML) Dean – Enterprise Solutions & Academic Collaborations https://www.linkedin.com/in/ravi-shankar-70584b1/ [email protected] First Session , 29th AUG, 2023 Introduction – Global Agri : Challenges & Solutions Faculty Profile : Dr. Ravi Shankar INTRODUCTION • Background: PhD - Econometrics, Entrepreneur, Start-up Mentor, Social Impact Investor, Data Science Thought Leader, Senior Venture Partner • Current Focus: AI adoption in Business, Deep Learning, XAI, ML Operations, Design Thinking, OB & Strategy • Sectoral Exposure: multiple sectors straddling BFSI / Pharma / Manufacturing / Retail / Tech • Overall Experience : 30 Yrs. Applied Analytics for Business - Agriculture: Learning Journey Five sessions of 90 mins each First session Second session Third session Fourth session Fifth session Agri 4.0 & Agri Data AI-ML Applications in Agri – use cases IntroductionThe Global Agricultural Challenge & Solutions Agri & AgTech Indian Landscape AgTech & Agriculture in India STRUCTURE & Learning outcomes… The Agricultural Challenge https://youtu.be/jF07b7IDxus Solution https://youtu.be/tQ5jADa0DAs https://youtu.be/25sQYqviC28 Big Data in Agri https://youtu.be/GKfJU5rYHYo https://youtu.be/4r_IxShUQuA https://youtu.be/WKGQQXwZDQ4 https://youtu.be/Ks0lBuaqyjM 1. 2. 3. 4. 5. Global Agri Challenge & Solutions The intuition behind Big Data in Agri Data Science in Agri Challenges relating to application Real world examples Digitization in agriculture, precision agriculture, drones and IoT in farming. “Demonstration of ICT applications and Agri informatics” Speaker Profile : Dr. Ravi Shankar INTRODUCTION • Background: PhD –Agril Economics, Visiting Prof Ai-ML, Entrepreneur, Startup Mentor, Social Impact Investor, Data Science Thought Leader, Senior Venture Partner • Current Focus: AI adoption in Business, Deep Learning, XAI, ML Operations, Design Thinking, OB & Strategy • Sectoral Exposure: multiple sectors straddling BFSI / Pharma / Manufacturing / Retail / Tech • Overall Experience : 32 Yrs. 6 Structure • • • • Motivation Backdrop Challenges & solution Digitization of Indian Agriculture • IoT in Indian Agriculture • Use of Drones in Indian Agriculture • Satellite Imagery applications in Indian Agriculture • Natural Farming in India • Role of AgTech Startup’s • Summary "We Didn't Run Out Of Television Screens And Planes, We Ran Out Of Food." The best Interstellar quotes not only sum up the themes of the movie but are generally thoughtprovoking lines. Joseph Cooper (Matthew McConaughey) and his daughter Murph (Jessica Chastain) are only two of the highly intelligent characters that make up the core cast of the movie. With scientists, astronauts, and advanced robots as the main characters, Interstellar is bursting with smart dialogue about scientific and philosophical concepts. In this thought-provoking Interstellar quote, the school's principal (David Oyelowo) reminds Cooper why farmers have become more important than engineers in their world. https://youtu.be/qhW1HfSuPVQ Motivation • Relevance of digitalization in India Agriculture cannot be overstated • Indian Agriculture – facing unprecedented challenges • • • Growing population Climate change Resource constraints • • Data driven decisions Management efficiency • • • Reduce costs Minimize environmental impact Optimise farm operations • Integration of Digital Technologies & Data Management systems into farming practices & supply chains • Solution – Digitalization • Benefits - many • KSF The evolution of Indian Agriculture – From Scarcity to Optimisation…. Solution – Digitization / Agri 5.0 Paradigm Mid 21st Century – Quality, Safety & Optimisation Traditional farming methods face multiple challenges, including resource depletion and environmental degradation. But with the concept of digitization, the future of agriculture looks promising. Early 21st Century – Efficiency related challenges 1960’s & 70”s – Green Revolution – adequate quantities Late 19th & early 20th Century – Starvation & Supply side constraints Indian Agriculture: An Overview 1 Large Sector 2 Small Holdings Indian agriculture is one of Most farms in India are the largest agricultural small, with an average systems in the world, landholding size of less contributing significantly to than 2 hectares. These the country's GDP and small farmers struggle with employing nearly half of the a lack of resources, total workforce. unpredictable weather, and market volatility. 3 Diverse Crops India is home to a wide variety of crops, including rice, wheat, cotton, sugarcane, and many more. The country is also a major exporter of spices and fruits. Challenges Faced by Indian Farmers Unpredictable Weather Limited Access to Resources Indian farmers frequently face unpredictable Many small farmers don't have access to weather patterns and natural disasters that modern farming technology, adequate can damage crops. irrigation, or quality seeds or fertilizers. Market Volatility Poor Infrastructure Fluctuating market prices and lack of market India's rural infrastructure, including roads information often leave farmers without and electricity, is often inadequate, making it access to fair prices for their crops. difficult for farmers to transport goods to markets and access modern technologies. THE AGRICULTURAL CHALLENGE A sobering statistic: Food production must increase 50-70 % by 2050 to feed a projected population of 9 Billion people “The End of Plenty” WORLD POPULATION Alexandratos N, Bruinsma J.2012.World agriculture towards 2030/2050, the 2012 revision. ESA Working Paper No. 12-03, June2012. Rome: Food and Agriculture Organization of the United Nations (FAO) AGRICULTURAL DEMAND POPULATION MORE PEOPLE CONSUMPTION HIGHER DEMAND / CAPITA GREEN REVOLUTION NORMAN BORLAUG IS CREDITED WITH SAVING 1 BILLION LIVES YIELD INCREASES Ray DK, Mueller ND, West PC, Foley JA. 2013.Yield Trends Are Insufficient to Double Global Crop Production by 2050. PLoS ONE 8(6), doi:10.1371/journal.pone.0066428. crop yield must I N C R EA S E * to meet demand by 2050 Ray DK, Mueller ND, West PC, Foley JA. 2013.Yield Trends Are Insufficient to Double Global Crop Production by 2050. PLoS ONE 8(6), doi:10.1371/journal.pone.0066428. IT’S POSSIBLE “I now say that the world has the technology… to feed on a sustainable basis a population of 10 billion people.” Normal Borlaug DATA SCIENCE MEETS AGRICULTURE NEXT REVOLUTION ? GREEN REVOLUTION 1960 – BIOTECH REVOLUTION 1980 – GREEN DATA REVOLUTION 2010 – INTENSIFY Apply breeding, fertilization to increase yields. BIOTECH Marker assisted selection. OPTIMIZE Apply data science to optimize management. DATA SCIENCE How can it be built? Computer Science SCIENTIFIC DATA SCIENCE use software engineering to enable domain science maximizing use of data Domain Science What is important? Statistics How can predictions be made? Annually one-third Working withof crops are lost globally 27% Loss in crops due to imbalance or lack of nutrients Upto 40% Loss in crops due to pest and diseases 20% Loss in crops due to water stress 30% Loss in the farm income. 33% Land is already degraded and over 90% could become degraded by 2050. SOC is ~0.5%. 0.3 m/yr Declining water table in India. Source: FAO. Global food losses and food waste – Extent, causes and prevention. Existing solutions are not reaching farmers Working with Precise ● ● ● ● Drone IoT sensors Agri expert Laboratory Scalable Required solution which is precise and scalable ● ● Standard package of practices Local retailers ● ● Source: Lowder, S. K., Skoet, J., & Raney, T. (2016). The number, size, and distribution of farms, smallholder farms, and family farms worldwide. World Development, 87, 16-29. Globally, about 80% of the farms are small and marginal. Currently available solution are either precise or scalable Digital Agriculture : Integration of cutting edge technologies (AI-ML | Big Data Analytics | IoT) Identify the problem Data collection Data analysis Solution implementation Benefits of Digitization Efficient resource management Improved yields and quality Digitization allows for precision agriculture, Digital tools such as drones and sensors can maximizing the use of resources and monitor crop health and alert farmers to minimizing waste. potential issues. Greater data visibility Cost-effective Digital records make it easier to track crop Digitization reduces the need for manual growth, soil moisture, and more. labor, helping farmers save money and time. Source: Niti Aayog – Role of Digital and AI Technologies in Indian Agriculture: by Tanay Mahindru Challenges with AI/Digital ● Clean Data Sets ● Lack of data sets – example – pest on leaves ● Diverse data sets between public, private and govt departments ● Lack of normalized data sets ● Each state has different language and units, which makes it harder for data normalization. ● Acceptability & adoption ● Who will pay for this, given so many small Farmers ● Interoperable standards needed in Agristack being developed by various players The Potential of Digital Agriculture 1 Climate-Smart Agriculture 2 The Internet of Things (IoT) Digital tools can assist with weather Sensors enable farmers to receive forecasting and mitigation of climate- up-to-date information about crop related risks. health and weather conditions, improving decision-making. 3 Farm Management Software 4 Fintech for Farmers Software like FarmERP helps Startups like FarmFundr and Clover, farmers optimize resources, manage leverage fintech to provide loans to inventory, and improve traceability farmers for seeds, fertilizers, and across the supply chain. equipment, helping them overcome financial barriers. Areas that AI can solve - Full Supply Chain Summary View Easy Access to Credit Insurance payout linked to more accurate yield Early Pest Detection Early Locust Detection Drone based pest spray and monitoring Precision Farming Farm Monitoring and Price/Yield Advisory Rental of equipment Farm Inputs Financing Which crop to sow Advisory Knowledge Availability on Social media Seed Traceability Farming Selling and Distribution Product Traceability Aggregator Marketplaces for efficient price discovery Optimal Harvesting Calendar using Price forecast Farmer Produce Organizations Farm-to-Fork Agri Value Chain Sowing Pre-Sowing ● ● ● Farm Inputs ● Land Preparation ● Soil Testing Machinery Management Sowing Advisory Crop Cultivation ● Farm Monitoring ● Transportation ● Weather Advisory ● ● Disease & Pest Management Yield Management ● Storage Management ● Packaging ● Traceability ● Irrigation and fertigation management Post Harvest Harvest ● Market Linkage 32 © Wolkus Technology Solutions Private Limited 2020 Nutritional Security An Existential Threat Malthusian Necessities of Life Food, Shelter, Fiber, Fuel > 9 billion Path Forward • Transformative discoveries – Internet of Agricultural Things – Big Data • • • • • • 21st Century Extension Farming systems Education Policies, regulation, marketing Human dimensions Communications #InternetOfAgThings In a world where population is expected to grow significantly, natural resources become scarcer and climate changes dramatically, we need to produce much more with much less. 21st Century Farm Supply Chain • • • • • • • • • • On Farm Production Soil Health, Water, Nutrients Pests/Control Energy Traceability and Tracking Supply Chain Management Processing Inspection Transportation Storage • • • • • • • • • • Retailers Inventory Access Smart Refrigerators Food Safety Ripeness Shrink Wrap Waste Smart Services Etc. Food Waste and Food Loss ➢ Double food production in 40 years ➢ Cut loss/waste by half? ➢ Impact climate change Adapted from Bonn2011 Nexus Conference: http://www.water-energy-food.org/en/news/view 255/understanding-the-nexus.html http://phys.org/news/2014-06-date-contributes-food.html Precision Foods • Individual genome, epigenome, microbiome • Plant/animal genome, epigenome, microbiome • Wearable sensors – FitBit, Apple Watch – Athos, Hexoskin, Gymi – Google contact lens • Food analysis • Lifestyle • Behavior Revolutionizing Agriculture with ICT and Agri Informatics Discover how ICT and advanced informatics systems are transforming agriculture and optimizing food production across the globe & in INDIA Digitization in Agriculture: Transforming the Way We Farm Benefits 🌱 Challenges 🌾 Access to real-time data, better decision From high implementation costs to data making, and increased efficiency are just management challenges, we explore the some of the benefits of digitization in main hurdles of digitization in agriculture agriculture. and how to overcome them. Digital Indian Agriculture • 2022 Union Budget: • • • Emphasis on Kisan Drones Chemical free Natural Farming PPP for delivery of digital & high tech services to farmers • 2023 Union Budget : Focus on Digital Agriculture • • • • • • • Digital Agriculture Mission – with a primary focus on digital infrastructure, digital extension services, digital finance and digital research and development India Digital Ecosystem of Agriculture (IDEA) Support in digital infrastructure development and improving its adoption by farmers Promoting data-driven ecosystem and digital services through Agri Stack Incentivizing capital investments at farmgate levels to boost the MSME segment and promote rural entrepreneurship Promoting Agritech Startups Focus on output linkage through e-commerce Precision Agriculture: A New Paradigm for Sustainable Farming Definition Examples Precision agriculture is a science-based From soil moisture sensors to drones, we approach to farming that uses technology to showcase the latest and most innovative optimize crop yield while minimizing waste. technologies driving precision agriculture. Drones in Agriculture: Applications and Benefits 1 Farm Planning 📝 2 Pest Control 🐛 Drones offer valuable Drones can be equipped insights into crop health, with targeted spraying which helps farmers make systems to efficiently better decisions to deliver pesticides and optimize crop production. herbicides to crops while reducing waste. 3 Mapping 🗺️ Drones can be used to create highly accurate maps of crop fields, providing farm managers with invaluable information on soil properties, groundwater, and weather patterns. Internet of Things (IoT) in Natural Farming: Connecting Our Farms to the Future Implementation Challenges Future Trends From technical difficulties to data From the use of big data and AI to management issues, we explore the development of new sensors the main challenges of and monitoring technologies, we implementing IoT in natural explore the exciting future of IoT farming. in natural farming. 1 2 3 Role and Impact Case Studies IoT is transforming natural farming We showcase successful by providing real-time data and implementation stories from insights, allowing farmers to around the world, demonstrating optimize yield and reduce costs. how IoT can make a difference in natural farming. 4 Case Studies of Successful Implementation: How Leading Farms Are Harnessing the Power of Technology to Boost Yield and Efficiency FarmBot 🤖 BlueRiver Technology 💦 John Deere 🔍 capabilities, FarmBot has BlueRiver has created a and data analytics solutions, helped farmers reduce labor breakthrough technology that John Deere has become a costs while boosting crop uses computer vision and AI leading name in precision yield through highly precise to deliver real-time agriculture, helping farmers planting and watering. recommendations on crop optimize yield and reduce management, boosting yield costs with ease. With its fully automated and reducing water usage. With its advanced machinery Conclusion: Key Takeaways and Future Trends What We've Learned 🧠 Looking Ahead 🔮 We've seen how ICT, agri informatics, and We explore the exciting future of agriculture and IoT are transforming agriculture and natural natural farming, from the increased use of big farming, bringing numerous benefits and new data to the development of new technologies and opportunities for growth. applications. Use of Drones in Indian Agriculture Indian agriculture is facing unprecedented challenges. This section will explore how drones can be used to address these challenges and revolutionize the industry. Drones: The Capabilities and Benefits 1 3 Aerial Imagery 2 Precision Agriculture Drones capture high-resolution images Drones can be equipped with sensors that allow farmers to monitor crop for better data collection, allowing health, growth patterns, and identify farmers to optimize their use of problem areas. resources and increase yields. Faster and Cost-Efficient 4 Safety Drones can cover large areas quickly Drones eliminate the need for farmers and accurately where traditional to navigate tough terrain or handle methods would be slower and more chemicals, improving safety and expensive. reducing accidents. Case Studies: Successful Drone Use in Indian Agriculture Rice Farming Drones were used to capture real-time data on crop health, leading to a 30% reduction in water usage and increased yields. Tea Plantations Drones helped assess soil moisture, leading to precise irrigation, which increased productivity and better tea quality. Cotton Farming Drones helped identify pest infestations, leading to reduced chemical usage, cost savings, and higher quality yields. Wheat Farming Drones equipped with multispectral cameras collected data on soil variations, resulting in targeted fertilization and increased yields. Regulations and Permissions for Drone Use in Agriculture 1 Current Regulations 2 Challenges 3 Future Outlook Operators face The government is India's Ministry of Civil bureaucratic delays and taking steps to Aviation has issued prohibitive costs in streamline regulations, guidelines for drone obtaining necessary encourage innovation, operations, requiring permits. and increase users to obtain the accessibility to drone necessary permissions technology. and licenses. Costs and ROI of Drone Implementation Drone Costs ROI Entry-level drones for Drones can help farmers agricultural use start at save on resources, reduce While drone around INR 1.5 lakhs, with costs, and increase yields, implementation in high-end models that can resulting in an estimated agriculture requires an cost over INR 40 lakhs ROI of 13% to 150% initial investment, it is a depending on the size and depending on the long-term investment and capabilities. application and can drive substantial implementation strategy. economic growth in the Long-term Investment industry. The Future of Drone Use in Indian Agriculture 1 Enhancing Data Analytics Improved analytical capabilities will lead to enhanced decision-making and more efficient use of resources. 2 Solar-Powered Drones With advances in solar-power technology, drones will become more sustainable and reliable for operations over longer periods. 3 Collaboration Farmers, drone manufacturers, and software industries will need to work together to continue improving technology and ensure successful implementation in agriculture. Conclusion Drones offer immense potential for Indian agriculture It's a revolution It is time for the Indian agriculture sector to the industry. In conclusion, drones are a game embrace the change and adopt drone changer in Indian Agriculture!!! technology for better productivity and ecological benefits. Drone technology can help address the challenges of Indian agriculture and transform IoT in Indian Agriculture Discover how IoT is revolutionizing the agriculture industry in India and helping farmers tackle the challenges they face. Introduction to IoT Sensors Automation IoT devices in agriculture use sensors to gather data on soil moisture, temperature, weather, crop growth, and other factors. IoT devices can automate tasks such as irrigation, fertilizer application, pest control, and harvesting, freeing up farmers' time and reducing costs. Connectivity IoT devices can be connected to smartphones, computers, and other devices, allowing farmers to monitor and control their farms remotely and access important data in real time. Cloud Storage IoT devices can store data on cloud platforms, providing farmers with accurate and detailed information that can be used for future planning and decision-making. How IoT is Being Implemented in Indian Agriculture Smart Irrigation Real-Time Weather Data IoT devices can be used to IoT devices can provide farmers automate irrigation systems and with real-time weather data that control water usage, improving can be used to make informed crop yields and reducing water decisions about crop waste. management. 1 2 3 Soil Monitoring Pest Control IoT sensors are used to monitor IoT devices can be used to detect soil moisture levels and nutrient and control pests, reducing the levels, allowing farmers to make need for harmful chemicals and data-driven decisions about improving crop quality. irrigation and fertilization. 4 Benefits of Using IoT in Agriculture 1 3 Increased Efficiency 2 Better Decision Making IoT devices can automate many IoT devices provide farmers with farming processes, reducing the need detailed and accurate data that can be for manual labor and increasing used to make informed decisions efficiency. about crop management. Cost Savings 4 Improved Sustainability By reducing water and chemical IoT devices can help farmers reduce usage, automating tasks, and water waste, harmful chemical usage, improving crop yields, IoT can help and other unsustainable farming farmers save money and increase practices, leading to a more profits. sustainable approach to agriculture. Examples of Successful IoT Projects in Indian Agriculture Gramophone eKutir Fasal Gramophone is an IoT eKutir is an IoT platform that Fasal is an IoT device that platform that provides provides farmers with access provides farmers with data on farmers with access to real- to certified seeds, fertilizer, temperature, humidity, and time weather data, soil and credit, as well as soil moisture, allowing them analysis, and crop advisory marketing channels for their to make data-driven services through their crops. decisions about irrigation, smartphones. fertilization, and pest management. Conclusion and Future of IoT in Indian Agriculture Overall, IoT has the potential to revolutionize Indian agriculture by increasing efficiency, reducing costs, and improving sustainability. As more farmers begin to adopt these technologies, we can expect to see even more innovative IoT projects emerge in the coming years. Satellite Imagery in Indian Agriculture Discover how Satellite Imagery is revolutionizing the agriculture industry in India and helping farmers tackle the challenges they face. Solution: mobile app based on the satellite remote sensing technology Working with Providing real time information based on the satellite remote technology for which patent is pending. Soil nutrient report (NPK, pH and SoC) Fertilizer dosage recommendation Source: Sat2Farm’s product information Crop health monitoring Pest & disease forewarning & Diagnosis ● ● ● ● ● Geotagging and instant information No hardware required No capex required No needs to carry samples anywhere Recommendations on optimal: ○ ○ ○ Pesticide/insecticide Fertilisers Irrigation Sat2Farm: with Easy to use for farmers Working Farmer geotag’s his farm Server processes the satellite image Using walk/draw option in App Using cutting edge technology including AI/Ml 01 02 Farmer registers in App Get the login credentials of app 03 Satyukt team receives satellite images for marked area The raw historical/present data is fetched Source: Satyukt, Sat2Farm app 04 05 Farmer get farm related information Results Working with 60 % 45 % 25 % Increase in yield Reduction in the cost Increase in yield Saved the crop by pest/diz̄ sease By applying optimal amount of fertilizers Source: FAO. Personal interactions with the farmers Through finding the appropriate problem Results Working with The farmer from the Nagarkurnool district in the state of Telangana was able to increase his crop yield by 60%. The farmer from the Ramnagar district in the state of Karnataka saved the cost on fertlizer by 45% through the rapid soil testing and dosage recommendations. The farmer from the Bijnor district in the state of Uttar Pradesh was able to find the appropriate problem which was the lack of nitrogen rather than any pest or disease as contemplated by many locally. He was able to increase the yield by 25%. Source: FAO. Personal interactions with the farmers Summary andwith way forward Working ➢ It is possible to estimate information from satellite remote sensing for the management of: – – – Water resources Nutrients Pest/diseases ➢ Currently used solutions are useful only for a fraction of farmers. ➢ It is possible to bring precision farming for every farmer. ➢ It is possible to combine modelling and observations to solve agriculture problems. Source: Tomer, S. K. et. al (2016) AgTech Startups: Best Suited to Turbocharge Digital Agri in India AgTech, a sub-sector of Agritech, focuses on developing technology to solve agriculture challenges and improve the efficiency of farming processes. 1 Domain Expertise 🌾 AgTech startups have vast knowledge and experience in the agriculture domain, which enables them to innovate efficiently. 3 Innovation 💡 2 Up-to-date with Technologies 🔬 AgTech startups are always updated with the latest technological advancements and tailor it to the Agriculture Industry needs. AgTech startups are known to create breakthrough technologies and provide unique solutions to the challenges faced by farmers. Conclusion The digitization of natural farming has the potential not only to revolutionize agriculture but to make it more sustainable and eco-friendly. The successful integration of digital tools with natural farming methods has the potential to improve yields, increase resource efficiency, and provide more nutritious food for the world’s population. Grow More, Grow Better Up to 40% increase in yield. Up to 60% reduction in pesticide usage. Up to 50% reduction in water usage for irrigation. Reduction in usage of non renewable fuel resources. Accurate usage of fertilizers required. Forbes Asia 30 under 30 Improvement in soil health with less usage of fertilizers. #BigData in Agri CHALLENGES DATA CHALLENGES LEARNING CHALLENGES Spatio‐‐TemporalData Latent Features Sparse Data Curse of Dimensionality Missing Data Multi-‐task Learning Noisy Data EASY! RIGHT? unfortunately, no DATA POTENTIAL one season, one crop, one country YIELD MONITOR DATA 14B OBSERVATIONS REMOTE SENSING DATA 260B OBSERVATIONS WEATHER DATA 20B OBSERVATIONS FEATURE ENGINEERING LATENT SPACE Genetics, Environment, Practices Soil Processes Nutrient Processes Crop Processes Yield Zea mays (corn) LATENT FEATURE SPACE Environment, Genetics and Practices Engineered Features { Physical Processes Yield Outcome } Learned Features CAUSAL DESCRIPTION THE ROLE OF DEEP LEARNING FEATURE LEARNING yield genetics, environment and practices hidden layers physical models soil processes genetics, environment and practices Deep Neural Network nutrient processes crop processes yield Hierarchical Dimensionality Reduction SPATIAL DATA CONVOLUTIONAL DBN Hierarchical representation of spatial data High-dimensional, scalable visible layer Unsupervised hierarchical learning Lee, Honglak, et al. "Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations." Proceedings of the 26th Annual International Conference on Machine Learning. ACM, 2009. MULTI-TASK LEARNING y Hidden layers (latent features) shared across tasks Multi-task informs latent features w Hidden Layers DEEP NEURAL NETWORK genetics, environment and practices deep neural network in multi-task setting Deng, Li, Geoffrey Hinton, and Brian Kingsbury. "New types of deep neural network learning for speech recognition and related applications: An overview." Acoustics, Speech and Signal Processing (ICASSP), 2013 IEEEInternational Conference on. IEEE, 2013. MISSING DATA DEEP BELIEF NETWORK Greedy layer-wise training algorithm Robust to noisy inputs Generative process (MRF) Alternating Gibbs sampling in deep belief network Hinton, Geoffrey E., Simon Osindero, and Yee-Whye Teh. "A fast learning algorithm for deep belief nets." Neural computation 18.7 (2006): 1527-1554. OTHER APPLICATIONS Crop identification Practice classifications Disease detection Remote sensing Image segmentation / clustering Nutrient deficiency detection Cloud detection Environment classification additional applications of deep-learning in agriculture POSSIBILITIES PROTECT & IMPROVE REDUCE RISK INCREASE YIELDS Goal: optimize global food production PROTECT & IMPROVE The Climate Corporation aims to help farmers around the world protect and improve their farming operations & profitability. TWI Yield Potential TotalWeatherInsurance:parametric supplemental crop insurance product Yield Expectation TWI PROTECT Government subsidized loss-adjusted insurance program, integrated risk management Profitability GROWER APPLICATIONS MP MP GROWER APPLICATIONS Collection of grower management advisors using agronomic and climatological models IMPROVE Challenges with AI/Digital ● Clean Data Sets ● Lack of data sets – example – pest on leaves ● Diverse data sets between public, private and govt departments ● Lack of normalized data sets ● Each state has different language and units, which makes it harder for data normalization. ● Acceptability & adoption ● Who will pay for this, given so many small Farmers ● Interoperable standards needed in Agristack being developed by various players Big Data: Milieu • • • • • • • • • • • Analytics Informatics Evidence-Based Tools Meta-Analysis and Synthesis Complex Systems Computational Sciences Data Engineering Data Mining Cloud Computing Implementation and Evaluation Data Security and Cybersecurity • • • • • • • • • • • Predictive Modeling Data Visualization Decision Analytics Embedded Systems Machine Learning Multidimensional Data Network Science Sensor Networks Spatial Analytics Bandwidth Cyberphysical Systems Big Data: Opportunities Open Data is a powerful, evidence-‐based tool for long-‐ term sustainable development by improving economic opportunities for farmers and health of consumers. Open access to research, meta-‐ analysis, and open publication of data are vital resources for nutritional security. Big Data: Challenges • Ownership – Open Ag Technology Systems • • • • • • • • • Decision Support Tools Cost Bandwith Quality Curation Disambiguation Connectivity Cybersecurity Storage We are drowning in information, while starving for wisdom. The world henceforth will be run by synthesizers, people able to put together the right information at the right time, think critically about it, and make important choices wisely. E. O. Wilson, Entomologist, Author, Pulitzer Prize Winner ASSIGNMENT – A five page note on Wilson’s seminal contribution in creating a global database Big Data in sustainable agriculture • Big Data can support resilient agri-food systems under uncertain climate variability and change. • Big Data has a huge potential in the dry areas where resource use efficiency is below its actual potential • But they can only deliver if applied to Inclusive Farming Systems + + Better integration Better measurements Better modelling Resilient Systems Thank you