Evolutionary computational design of closed ecosystems using artificial gravity

One of the most important technologies to realize permanent space settlements is the development of self-sustaining controlled ecological life support systems (CELSS). This will require replication of independent self-contained subsets of Earth’s biosphere containing select flora and fauna under controlled conditions for eventual human life support. But are 100% closed ecosystems (with the exception of the exchange of radiation and information) beyond Earth possible? Could a series of controlled evolutionary experiments using machine learning be carried out on controlled ecosystems in space under variable gravity conditions to rapidly optimize the key variables needed to identify the smallest possible CELSS for long term human survival? Gregory Dorais, a research scientist at NASA Ames Research Center, thinks so and describes the strategy in a paper called An Evolutionary Computation System Design Concept for Developing Controlled Closed Ecosystems.

Dorais introduces his concept with a brief description of Closed EcoSystems (CESs) and early efforts by NASA to develop a CELSS for space settlement. Of particular concern are the challenges of putting humans in the equation. There are consequences related to the ratio between human biomass and non-human biomass in ecosystems. On Earth this ratio is low so the ecosystem can self-regulate compensating for imbalances. But in a space biosphere, this ratio in the life support system is comparatively huge leading to significant challenges in maintaining equilibrium. For example, the ISS needs frequent resupply of consumables by spacecraft to replenish losses in the life support system. Wastes that cannot be recycled are either incinerated in the Earth’s atmosphere or exhausted into space. A completely closed system that is self-sustaining has not yet been developed.

Dorais’ design concept for an experimental testbed can be used to explore the viability of different biomass ratios of various combinations of larger animal species and eventually humans. The system consists of a collection of independent CESs controlled and interconnected to generate data for machine learning toward optimizing long term viability. Gradually, the size of the animals in the CES can be increased evolving over time with the ultimate goal of human life support. To kick things off, an Orbiting Modular Artificial-Gravity Spacecraft (OMAGS) is proposed, with room for 24 CESs housed in a 150cm radius centrifuge with appropriate radiation shielding capable of testing the ecosystems under different fractional gravity conditions. The spacecraft is envisioned to be placed in an elliptical orbit in cis-lunar space.

The OMAGS spacecraft has been sized to fit in a SpaceX Falcon Heavy payload fairing.

A NASA patent and tech transfer fact sheet entitled Closed Ecological System Network Data Collection, Analysis, Control, and Optimization System has been issued for this innovation under the NASA Technology Transfer Program.

In a related presentation delivered in November 2018, Dorais says “Once CESs are demonstrated to reliably persist in space, within specified gravity and radiation limits, it is a small step for similar CESs to persist just about anywhere in space (Earth orbit, Moon, Mars, Earth-Mars cycler orbit, asteroids, …) enabling life to permanently extend beyond Earth and grow exponentially.”