[ad_1]
The days when you could just grow a basil plant from a seed by placing it on your window sill and watering it regularly are over – there is no reason anymore now that hydroponic "e-agriculture" optimized for machine learning produces a superior plant with more robust flavors. The future of pesto is here.
This research was not born of a desire to improve sauces. This is a study of MIT's Media Lab and the University of Texas at Austin that aimed to understand how to improve and automate agriculture.
In the study, published today in PLOS ONE, the question that was asked was whether a growing environment could find and execute a growth strategy leading to a given goal – in this case, the basil flavored stronger.
This task has many variables to change – soil type, plant characteristics, frequency and volume of watering, lighting, etc. and a measurable result: the concentration of molecules producing aromas. This means that it is ideally suited to a machine learning model that, from this variety of inputs, can predict which one will produce the best return.
"We are really interested in building networked tools that can take the experience of a plant, its phenotype, all the constraints it meets and its genetics, and digitize this to enable us to understand the plant-environment interaction, "explained Caleb Harper of MIT. in a press release. The better you understand these interactions, the better you can design the plant's life cycle, perhaps increasing yield, improving flavor or reducing waste.
In this case, the team limited the machine learning model to the analysis and switching of the type and duration of the light felt by the plants, in order to increase the concentration of aromas.
A first set of nine plants had hand-designed light regimes based on prior knowledge of what basil usually likes. The plants were harvested and analyzed. Then, a simple model was used to develop similar but slightly modified treatment regimens, taking into account the results of the first cycle. Then, a third, more sophisticated model was created from the data, which gave it a considerably greater leeway to recommend changes to the environment.
To the surprise of the researchers, the model recommended an extremely extreme measure: keep the UV lamps on the plant on 24/7.
Naturally, this is not how basil grows in the wild, because, as you may know, there are few places where the sun shines all day. and all night long. And the Arctic and Antarctic, while fascinating ecosystems, are not known for their herbs and their spices.
Nevertheless, the "recipe" of keeping the lights on was followed (it was an experience, after all) and, incredibly, this resulted in a massive increase in aroma molecules, doubling the amount found in the control plants.
"You could not discover that otherwise," said co-author John of Parra. "Unless you're in Antarctica, there's no 24-hour photoperiod to test in the real world. Artificial circumstances were needed to discover this. "
But if a tasty basil is a positive result, that's not really the problem. The team is happier that the method has generated good data, thus validating the platform and the software used.
"You can see this document as the starting point for many applications that can be applied, and it's an exhibition of the power tools we've built so far," Parra said. "With systems like ours, we can dramatically increase the amount of knowledge that can be acquired much more quickly."
If we want to feed the world, it will not happen with waves of amber cereals, that is to say with traditional farming methods. Vertical, hydroponic, computer optimized – we will need all these advances and more to bring food production to the 21st century.
[ad_2]
Source link