Computer Vision and Sequential Decision Making for Plant Cultivation in Bioregenerative Life Support Systems

Bioregenerative life support systems (BLSS) are being considered for future long-duration manned space travel, as supplies from Earth are too expensive, if not impossible. Moreover, environments suitable for agriculture will also be difficult to access on Earth, mainly due to climate change and resource depletion. Both on Earth and in space, it is therefore necessary to find solutions for growing plants, which can benefit from robotics, advanced automation and machine learning, according to the scientific literature.

This internship project in artificial intelligence applied to precision agriculture aims to contribute to the development of methods for maximising long-term production and minimising resource consumption.

It focuses on optimising:

1. the estimation of the state of plants and their environment using machine learning and computer vision algorithms to monitor their growth,

2. sequential decisions using reinforcement learning algorithms to calculate autonomous cultivation strategies that are efficient and economical in the long term.