AlgAIe - Optimering av algodling med AI för att avsevärt minska metanemission från nötkreatur

Purpose and goal:

This project brought together Volta Greentech, Chalmers Industriteknik, IVL Swedish Environmental Research Institute, Swedish University of Agricultural Sciences and Tre Bönder to explore how AI-based optimization methods can increase the yield and quality of macroalgae (Asparagopsis) in large-scale cultivation systems with the mission to combat climate change. A biomass sensor was also developed to increase harvest efficiency and in-vitro and in-vivo studies were conducted to understand the impact growing conditions of algae have on their methane reduction potential.

Expected results and effects:

A biomass sensor using light was developed that can read and track the growth of algae in the tanks, providing valuable information about when to harvest. Indications are that you can measure biomass with very high accuracy, 3 percent error, in a cheap and simple way. An important AI observation was that the original tanks used at VGT production facility were too tall and did not let enough light through to the bottom.

Therefore, new shallower raceway tanks were developed which are now in use. In-vitro and in-vivo testing demonstrated methane emission reductions of around 80 percent. Approach and implementation:

VGT performed cultivation tests where different parameters were considered (e.g temperature, pH, light intensity), the data generated from these experiments were then used in the ML models. The AI team looked at five areas for model development: 1) Optimization for seaweed cultivation parameters, 2) Nutrient consumption analysis, 3) Culture depth optimization, 4) Relationship of dissolved oxygen and biomass, 5) Biomass forecasting.

IVL tested three different techniques to measure biomass, made a prototype and validated the system in several growth cycle of the algae.