Simulating Polyculture Farming to Tune Automation Policies for Plant Diversity and Precision Irrigation

Abstract

Polyculture farming, where multiple crop species are grown simultaneously, has potential to reduce pesticide and water usage, while improving the utilization of soil nutrients. However, it is much harder to automate than monoculture. As a first step toward developing automation control policies for polyculture farming, we present AlphaGardenSim, a fast, first order, open-access simulator that integrates single plant growth models with inter-plant dynamics, including light and water competition between plants in close proximity. The simulator approximates growth in a real greenhouse garden at 9, 000X the speed of natural growth, allowing for policy parameter tuning. We present an analytic automation policy that in simulation reduced water use and achieved high coverage and plant diversity compared with other policies, even in the presence of invasive species. Code and supplementary material can be found at https://github.com/BerkeleyAutomation/AlphaGarden.