 |
Activities
Advanced System Modeling and Control of Bioregenerative Life Support
It may seem strange to think of the first manned spaceship to Mars as a flying farm. But consider how difficult it would be to carry all the food a crew would need for what might be a three-year round-trip voyage.
Much better to grow your food as you go. And if your crops also help you to recycle your water and air, you've got a nice little ecosystem to support you through your journey.
Of course, a spacecraft isn't exactly the wide open spaces of Iowa. So each plant will really have to pull its weight, consuming resources in the most efficient way possible to yield the maximum amount of food. And our spacefaring farmers will have to cultivate their crops in such a way that humans and plants can live in harmony, sharing the ship's precious supply of air and water to their mutual benefit.
Developing that super-efficient system requires an intimate understanding of plant growth beyond what now exists. So AEMC is developing a computer algorithm, modeled after the human brain, to learn how potential space crops grow, and ultimately to control all aspects of their life cycles.
Green factories
It may be a coincidence that "plant" is a synonym for "factory." But a green plant really is a factory, and a complex one at that. Through photosynthesis, it converts light energy into chemical energy, which becomes available to us and our fellow creatures as the food that sustains nearly all life on Earth. In the process, the plant removes carbon dioxide from the atmosphere, replenishes the air with oxygen, and releases fresh water from its leaves.
If AEMC's algorithm is to sit at the controls of this green machine and keep it humming at peak efficiency, it must first learn exactly how the controls work.
The extent to which a plant produces food, oxygen, and water varies with the kind of plant, its age, and assorted environmental conditions. The relationships are complex, nonlinear, and dynamic - just the sort of fuzzy input/output for which an artificial neural network is suited!
The Neural Network
A neural network is an algorithm with some of the learning characteristics of a biological brain. Our brains take the experiences of our five senses and draw conclusions that help us cope with our changeable world. Similarly, AEMC's artificial neural network takes in data about how plants perform under specific sets of conditions, and infers how various other combinations of conditions would affect that behavior.
 |
| Soybean plants in a growth chamber at Rutgers University. |
|
The system gets its raw data from an airtight hydroponic growth chamber in which plants such as soybeans, wheat, and lettuce grow at assorted tightly controlled levels of light intensity, air temperature, relative humidity, and atmospheric carbon dioxide concentration. At each magnitude of these environmental factors, the chamber records the plants' corresponding photosynthesis rate, transpiration rate (release of water), and allocation (the proportion of its biomass that is edible).
The neural network draws relationships between the plants' performance on one hand, and the kind of plant, age, and environmental factors on the other, and interpolates how the plants would function at intervening levels of those variables.
For example, if the neural network knows how much oxygen, food, and water the plants produced at 22�C, 26�C, and 30�C, it can estimate how much they would produce at temperatures between 22� and 26�, and between 26� and 30�. The real trick comes when it juggles all the variables at once!
Ultimately, the neural network will know how the plants behave under all conditions they are likely to experience. And it will be able to adjust their environments to maintain their performance at optimum levels.
If the crew aboard a future mission to Mars exercises strenuously - and therefore breathes harder, radiates more heat, and sweats more than usual - the neural network operating their hydroponic farm will know just how to adjust the crop's light levels, air temperature and humidity to compensate for the extra carbon dioxide, heat, and moisture the crew puts into their atmosphere.
Benefits to Earth
Earthbound farmers stand to gain as well from AEMC's work in this field. They, too, want maximum crop yield with minimum demand on resources. And they would benefit enormously from any new techniques that might help their crops survive droughts, frosts, and other vagaries of weather. AEMC's algorithms, and the information it is uncovering about the physiological processes of plants, may enable them to grow crops more efficiently and increase overall production of food on our planet.
For more information about life support, visit http://advlifesupport.jsc.nasa.gov/.
|
