Artificial intelligence (AI) has smoothly penetrated in a number of monitoring and control applications including the agriculture. However, research efforts towards low-power sensing devices with fully-functional AI on board are still fragmented. In this work, we present an embedded system enriched with the AI ensuring the continuous analysis and in-situ prediction of the growth dynamics of plant leaves. The embedded solution is grounded on a low-power embedded sensing system with a Graphics Processing Unit (GPU) and is able to run the neural networks-based AI on board. We use a Recurrent Neural Network (RNN) called the Long-Short Term Memory network (LSTM) as a core of AI in our system. The proposed approach guarantees the system autonomous operation for 180 days using a standard Li-ion battery. We rely on the state-of-the-art mobile graphical chips for ‘smart’ analysis and control of autonomous devices. This pilot study opens up wide vista for a variety of intelligent monitoring applications, especially in the agriculture domain. Also, we share with the research community the Tomato Growth dataset.

Human space exploration needs stable life support systems for the supply of oxygen, water and food for each human explorer due to long term missions. The most promising approach for building stable life support systems is a combination of physico-chemical and biological systems. These hybrid systems combine the reliability of physico-chemical and the sustainability of biological life support systems. Also the disadvantages, which are the finite resources of physico-chemical and the imperfect reliability of biological systems, are mutually balanced. To improve the reliability of biological life support systems, a combination of different biological systems may stabilize the whole approach during long term operations. The satellite mission Eu:CROPIS (Euglena gracilis: Combined Regenerative Organic-food Production In Space) is a testbed for investigating the behavior of combined biological life support systems under the influence of altered gravity, here, Lunar and Martian gravity. The core systems are a biological trickle filter for processing urine into a fertilizer solution via nitrification and Euglena gracilis, a photosynthetic protist which is able to produce oxygen and biomass while protecting the whole system against high ammonia concentrations.

Planetary habitation requires technology to maintain natural microbial processes, which make nutrients from biowaste available for plant cultivation. This study describes a 646 day experiment, in which trickling filters were monitored for their ability to mineralize nitrogen when loaded with artificial urine solutions of different concentrations (40, 60, 80 and 100% v/v). Former studies have indicated that increasing urine concentrations slow nitrogen conversion rates and induce growing instability. In the current experiment, nitrogen conversion rates, measured as nitrate production/day, did not differ between concentration levels and increasing instability was not observed. Instead, the buffering capacity of the mussel shells added as buffer system (∼75% calcium carbonate) increased with increasing concentrations of synthetic urine possibly due to the higher phosphate content. The intensified precipitation of calcium phosphates seems to promote carbonate dissolution leading to improved buffering. For space applications, the precipitation of calcium phosphates is not desirable as for the phosphate to be available to the plants the precipitate must be treated with hazardous substances. With regard to terrestrial agriculture the process-integrated phosphate precipitation is a possibility to separate the macronutrients nitrogen and phosphate without addition of other chemicals. Thus, the described process offers a simple and cost-effective approach to fertilizer production from biogenic residues like slurry.

The reutilization of wastewater is a key issue with regard to long-term space missions and planetary habitation. This study reports the design, test runs and microbiological analyses of a fixed bed biofiltration system which applies pumice grain (16-25 mm grain size, 90 m(2)/m(3) active surface) as matrix and calcium carbonate as buffer. For activation, the pumice was inoculated with garden soil known to contain a diverse community of microorganisms, thus enabling the filtration system to potentially degrade all kinds of organic matter. Current experiments over 194 days with diluted synthetic urine (7% and 20%) showed that the 7% filter units produced nitrate slowly but steadily (max. 2191 mg NO3-N/day). In the 20% units nitrate production was slower and less stable (max. 1411 mg NO3-N/day). 84% and 76% of the contained nitrogen was converted into nitrate. The low conversion rate is assumed to be due to the high flow rate, which keeps the biofilm on the pumice thin. At the same time the thin biofilm seems to prevent the activity of denitrifiers implicating the existence of a trade off between rate and the amount of nitrogen loss. Microbiological analyses identified a comparatively low number of species (26 in the filter material, 12 in the filtrate) indicating that urine serves as a strongly selective medium and filter units for the degradation of mixed feedstock have to be pre-conditioned on the intended substrates from the beginning.