Earth’s biosphere is sustained by its biological diversity, which forms an intricate network of biological, physical and chemical pathways. This network has many fail-safe redundant func-tions including buffer stocks of inert biomass, huge amounts of water and the large volume of gases in the atmosphere. By contrast, manmade habitats for human space exploration are closed ecosystems that represent only a trivial fraction of Earth’s biosphere. The employment of bio-regenerative processes complemented with physical-chemical tech-nologies is thought to have numerous advantages from the perspective of redundancy and reducing resupply mass for the sustained human presence in space or on other planetary surfaces. However, the combination of bio-regenerative processes, such as plant cultivation, with physical-chemical processes to form hybrid life support systems is challenging. Such systems are a concert of many interdependencies and interacting feedback loops, which are difficult to operate in a desired range of set points. Furthermore, the complexity of such sys-tems makes them vulnerable to perturbations. Applying system dynamics modelling to study hybrid life support systems is a promising ap-proach. System dynamics is a methodology used to study the dynamic behavior of complex systems and how such systems can be defended against, or made to benefit from, the per-turbations that fall upon them. This thesis describes the development of a system dynamics model to run exploratory simulations, which can lead to new insights into the complex behav-ior of hybrid life support systems. An improved understanding of the overall system behavior also helps to develop sustainable, reliable and resilient life support architectures for future human space exploration. A set of simulations with a hybrid life support system integrated into a Mars habitat has been executed and the results show a strong impact of space greenhouses on the life support sys-tem behavior and the different matter flows. It is also evident from the simulation results that a hybrid life support system can recover from a perturbation event in most cases without a fatal mission end. Recycling urine to produce a plant nutrient solution is a novel approach in further closing loops in space life support systems. Within this thesis, a number of experiments have been executed in order to determine the effectiveness of a urine-derived nutrient solution com-pared to a standard reference solution. The results show that in principle plants can be grown with a nutrient solution made of human urine, but that the yield is lower compared to the reference solution. However, the urine-derived solution might be tuned by adding small amounts of additional nutrients to remove the imbalance of certain elements. This way the nutrient salts supplied from Earth could be reduced.

The objective of the proposed research is to study the integration of humans and automation for the operation of regenerative life support systems (RLSS). RLSS combine physico-chemical and biological processes with the purpose of increasing the autonomy of space habitats and the life quality of their living organisms by properly reusing byproducts and regenerating consumable resources. However, these processes require energy and time to transform chemical compounds and organic wastes into nutrients, consumables, and edible products. Consequently, the maintenance of RLSS imposes a considerable workload on human operators. In addition, the uncertainties introduced by unintended chemical reactions promoted by material loop closure may create unexpected situations that, if unattended, could translate into performance deterioration, human errors, and failures. The availability of novel chemical and biological sensors together with computational resources enable the development of monitoring and automation systems to alleviate human workload, help avoid human error, and increase the overall reliability of these systems. This research aggregates sensor data and human-expert situation assessments to create a representation of their situation knowledge base (\gloss{skb}). The representation is used in a switched control approach to the automation of RLSS, for decision support, and human-automation coordination. The aggregation method consists of an optimization process based on particle swarms. The purpose of this work is to contribute to the methodological development of situation-oriented and user-centered design approaches to human-automation systems. Experiments and simulations are supported on the process of respiration in an aquatic habitat acting as a RLSS.