AI Case Study
Open Agriculture Initiative identifies optimal growing conditions and correlations for optimising basil flavour using machine learning
The Open Agriculture Initiative at MIT uses "food computers" to grow small amounts of crops in controlled environments equipped with a variety of sensors to collect data. Using machine learning, Sentient then has analysed the data to determine optimal growing conditions based on predicted outcomes with a goal to maximise some target - flavour for basil plants in this case. This reduces the number of experiments needed to be run and has resulted in discovering significant relationships between growing conditions that would not necessarily have been identified by researchers.
As posted on Medium: "In the fall of 2015, OpenAg had completed its prototypes of open-source, small and medium-scale hydroponic controlled-environment growth chambers that we call the Food Computers. Equipped with sensors, actuators and the open-source microcontrollers Arduino and Raspberry Pi, Food Computers are a great experimental tool because they allows us to monitor and adapt the environment of the plants growing inside, and then analyze which combinations of environmental or biochemical variables correlate to the best phenotypic response, or outcome... We chose basil (Ocimum basilicum) as a model organism, building on research that shows that flavor molecule concentrations in basil increase after exposure to stress (e.g. increased heat, salinity, UV light, water stress, chitosan). Imagine a group of expert farmers each come up with their ideal set of environmental conditions for growing very flavorful basil — light, soil, water, climate, etc.
We translate those conditions into actuatable code, run that code in a Food Computer, and test and correlate the levels of flavor (in this case, volatile molecules called monoterpenes, sesquiterpenes, and phenylpropenes) generated by each plant, each time its is grown. The results from those tests inform the next round of hypotheses for what combinations of environmental conditions produce the most flavorful basil, and the process continues, evolving into an optimized climate recipe as the generations proceed, and without the limit of conventional seasons."
From Medium: "In the agriculture industry today, there is no open database of “optimal” growing conditions with correlated outcomes for cultivating food. Historically, farmers relied on their own instinct, oral histories, and expertise to come up with specific growing conditions and nutrients for the right “recipe”. The goal was always the same — obtain the best crop outcomes for yield, flavor, and nutrition (to name a few). But you can’t have the perfect recipe when the variables keep changing, and that means each year and every crop are a gamble. Adding another layer of complexity, many crops only grow in certain climates and are expensive to transport. Both growing and getting certain foods in certain places in the world is simply too resource-intensive."
The model has discovered correlations both previously known by scientists, such as the inverse correlation between plant flavour and size, and those unknown, including better flavour with constant 24-hour light. Other "significant, nonlinear interactions between recipe variables" were uncovered according to Medium.
From the Sentient blog: "First, the group used Food Computers (those contained growing environments we mentioned above) to grow generations of a crop (in this case basil). Elements like UV light, salinity, heat, water stress, and more are controlled in these experiments and the crop yield is analyzed. Our scientific approach is based on optimizing a surrogate model of how the plant grows under different recipes. First, when the number of samples is small and only a few actuators (UV light types in the first instance, etc.) are varied, Gaussian processes can be used to predict what the outcome would be given a new recipe, and Bayesian optimization used to create suggestions for good recipes. Later on, as the dimensionality and number of samples grow, we use a neural network as the model and evolutionary algorithms (population-based or EDA) as the optimization method. The suggestions are then tested in the FCs and the containers, and the results used to train the model further."
From Medium: "Food Computers generate roughly three-million points of data, per plant, per growth cycle".