Sustainable intensification of global agriculture and food systems

In 2050, the global population is expected to increase by 2 billion people to 9.7 billion. Most of the population growth will occur in developing countries with accelerated economic growth and urbanization. This leads to a fast expansion of the middle class that is embracing a higher standard of living. These ongoing trends put unprecedented stress on food, energy, and water resources. Specifically, the global food production must increase by at least 70% by 2050. Global consumption of energy and water will also have to grow by at least 50% between now and 2050. Considering that current agriculture and food sectors already consume 70-80% of the world’s freshwater withdraws, 50% of habitable land, and 30% of energy produced, Considering that current agriculture and food sectors already consume 70-80% of the world’s freshwater withdraws, 50% of habitable land, and 30% of energy produced, the one question that is critical to the sustainable development of human society in the 21^st century and beyond is the following:

“How can we massively increase crop production without putting additional stress on land, energy, and water resources while ensuring sustainability for all?”

To tackle this grand challenge, CLAMS will conduct highly interdisciplinary and collaborative research to reexamine and renovate all stages during the R&D, production, distribution, processing, and consumption of agricultural products, thereby offering data-driven, systems-oriented solutions and insights to the sustainable intensification of global agriculture and food systems. We are excited to work closely with experts in agronomy, chemistry, environmental engineering, economics and finance, artificial intelligence, and operations research toward creating a sustainable food future by 2050 and beyond.

Stage 1

In the first stage, CLAMS is dedicated to accelerating innovations in seeds and crop protection R&D, which fundamentally impacts the landscape of all subsequent stages. In order to meet the customers’ ever growing needs, it is important for agriculture research organizations to keep innovating and expanding their R&D pipelines. However, the pace of innovation in both seeds and crop protection sectors among major agriculture companies has slowed down considerably since the early 2000s. This is mainly because the traditional innovation approach, which relies mostly on scientists’ hands-on experience and trial-and-error experimentation, can no longer catch up with the increasing complexity associated with innovating new cultivars with more favorable traits as well as synthesizing new crop protection actives that are structurally more complex. Therefore, agriculture companies must break out of the traditional innovation model and embrace digital transformation to accelerate the R&D process and reduce costs. Having recognized this, for the seeds sector, we are proposing an optimization approach for marker-assisted gene pyramiding to identify the optimal plant breeding schedule that leads to the desired individual with all favorable traits present while reducing the number of plants and crossings required by 50-70%. And for crop protection sector, we will develop state-of-the-art AI-driven crop protection active ingredient synthesis planning platform and build a smart robotic system to automatically perform organic synthesis process development.

Stage 2

In the second stage, CLAMS will address the sustainable manufacturing and supply chain of agrochemical actives and intermediates. The synthesis of one active typically involves 10 to 15 steps. Overall, hundreds of kilograms of chemicals, mostly solvents, are used to produce just one kilogram of active ingredient, resulting in an outstanding carbon footprint. Comprehensive life cycle assessment studies also suggested that most of the carbon footprint actually comes from the production and distribution of reagents and solvents. Therefore, to minimize energy consumption, total costs, and environmental impacts during agrochemical production and distribution, we will focus on two aspects. The first is to cut down the use of fresh chemicals by recovering and recycling used reagents and solvents using effective separation technologies. And in the second aspect, we are interested in designing a sustainable and resilient global supply chain network for agrochemical products and raw materials. This stochastic optimization framework will account for seasonal variations, as well as uncertainties that are unique to the nature of agrochemical industry, such as changes in environmental and safety regulations, the outsource of production activities to contract manufacturing organizations in third countries, and so on.

Stage 3

In the third stage, CLAMS will enable digital and precision farming using AI to help farmers make rational, informed decisions on which crop they should grow, what the expected crop yield is, and how to effectively monitor and improve crop health. Specifically, we will introduce Natural Language Processing (NLP) methods to extract and analyze agriculture news data and combining it with various supervised learning approaches to predict the price of grain futures traded at Chicago Mercantile Exchange. Next, CLAMS aims to introduce a new deep learning framework that analyzes publicly available multi-spectral satellite images, daily weather parameters, as well as historical yield data, to more accurately forecast crop yield at county level several months before harvest. Another problem of interest deals with optimal sensor placement for measuring field soil moisture level. The goal is to simultaneously identify the minimum number of soil moisture sensors as well as their optimal placement locations while considering potential sensor failure. In particular, we are interested in building an analogy between this problem and the game of Go, so that one can use machine learning approaches, as in the case of AlphaGo, to solve this historically challenging problem.

Stage 4

Lastly, CLAMS will look into promoting sustainability in food processing and consumption by identifying ways to minimize food waste, water and energy use, as well as environmental impacts during these final steps. In particular, we would like to study recent consumer trends in dietary habits and cooking preferences and understand their implications on agriculture and food systems, in order to promote sustainable end-user consumption behaviors among consumers and provide systematic guidance to upstream food producers on sustainability improvement. First, we are interested in the evolving food purchasing trend of switching from grocery meals to meal kits, a fast-expanding business that is experiencing a staggering 25% annual market size growth in the US. We will develop a life cycle optimization framework for both meal kits and traditional grocery meals to improve sustainability and bring down food and packaging waste for both business models. Another research interest lies in analyzing the environmental impacts as consumers shift their dietary preferences toward healthier diets, such as plant-based meat substitutes. While life cycle assessment studies exist for major plant-based meat brands, whether they were conducted in a systematic, unbiased manner remains unclear to many researchers. In response to these concerns and speculations, we will conduct the first rigorous hybrid LCA study on plant-based protein industry by incorporating appropriate system boundary, modeling assumptions, and function unit definitions.