Future Science: Supporting a mission to Mars

Conceptual illustration of a solar power driven biomanufactury on Mars. Illustration by Davian Ho.

As the concept of sending a long-term team of astronauts, scientists, and other specialists to Mars to setup a habitat becomes more realistic, NASA is asking scientists and engineers to look at solutions to some of the many problems that will confront the missions’s day to day operations.

One question being asked is: what is the most practical way to power future Mars missions? The seemingly simple question took UC Berkeley engineering students Anthony Abel and Aaron Berliner four years of hard work to figure out.

Most scientists and engineers who have thought about the logistics of living on the surface of the Red Planet have assumed that nuclear power is the best alternative, in large part because of its reliability and 24/7 operation. Solar power, on the other hand, must be stored for use at night, which on Mars lasts about the same length of time as on Earth. And on Mars, solar panels’ power production can be reduced by the omnipresent red dust that covers everything.

A new study, recently published in the journal Frontiers in Astronomy and Space Sciences, used a systems approach to compare the two technologies for a six-person extended mission to Mars involving a 480-day stay on the planet’s surface before returning to Earth.

“If humanity collectively decides that we want to go to Mars, this kind of systems-level approach is necessary to accomplish it safely and minimize costs in a way that is ethical. We want to have a clear-eyed comparison between options, whether we’re deciding which technologies to use, which locations to go to on Mars, how to go, and whom to bring,” said co-first author Anthony Abel, a graduate student in the Chemical and Biomolecular Engineering lab of Douglas Clark.

“Photovoltaic energy generation coupled to certain energy storage configurations in molecular hydrogen outperformed nuclear fusion reactors over 50% of the planet’s surface in our research,” said co-author Aaron Berliner, a bioengineering doctoral student in the Bioengineering Lab of Adam Arkin.

Their published findings showed that solar power was more economical toward the equator and nuclear power would be more efficient toward the poles. Read more about this research at chemistry.berkeley.edu/marspower.

Stefano Cestellos-Blanco will receive his Ph.D. this May from the College of Chemistry. He entered the College as a graduate student in the lab of Peidong Yang in 2016. He didn’t plan to figure out how to manufacture sugar on Mars as part of his research but what turned out to be an after-hours research project, stimulated by a NASA competition to spin sugar directly from carbon dioxide, is exactly what he and fellow scientists YiFan Li and Michael Ross accomplished.

CO2 is an abundant material on Mars. NASA was interested in a conversion process that would turn CO2 into sugar with the aim of feeding it to microbes that will make more complex materials, like food or drugs, for use by astronauts or settlers on Mars.

For the competition, the team demonstrated that the Formose reaction using formaldehyde, potentially from thermochemical CO2 fixation, and glycolaldehyde from CO2 electrosynthesis, generates sugars — from three-carbon sugars up to eight-carbon sugars — in about four hours, within the time frame specified in the competition.

“Converting CO2 directly to sugar is a pretty tall task that had never been demonstrated before, and NASA not only wanted you to demonstrate that you could do it, but also within a few hours, which is a relatively short amount of time,” Cestellos-Blanco said. “The individual parts of our process had been reported before, but no one knew that you could put them all together and essentially come up with a pathway to produce useful sugars from CO2.”

“We made a soup of sugars and have been able to identify which sugars those are. We also went ahead and fed the sugars to E. coli and grew them in cultures,” he continued. Read more about this project at chemistry.berkeley.edu/marssugar.

Scientist reviews biomanufactory operations on Mars. Illustration by Bryan Vector (Adobe Stock).

Associate Professor of Chemical and Biomolecular Engineering Ali Mesbah and graduate student George Makrygiorgos are participating in NASAs Center for Utilization of Biological Engineering in Space (CUBES*) program focused on research to advance learning-based predictive control methods with a view to optimize integrated biomanufacturing systems in real-time. 

Much of the complexity of space missions happens “behind the scenes”. Engineering systems capable of producing food, pharmaceuticals, and biomaterials must maintain their integrity to support the scientific exploration projects that analyze the habitat on Mars. Gaining efficiency in space biomanufacturing will be an especially important component to the survival of humans in this exotic environment far from earth.

Experimenting with different mission stage requirements for assembly, operation, timing, and productivity can help strengthen various optimal biomanufacturing system configurations. Graduate student George Makrygiorgos (advisors Ali Mesbah and Douglas Clark), is working on development of a data-driven modeling and optimization framework to address the high flexibility, scalability, and infrastructure minimization needs for just such an integrated biomanufacturing system.

According to George, “Relying on synthetic biology to support a crewed mission to Mars is a safety-critical endeavor. My research strives to explore decision-making strategies to increase biomanufacturing resilience so that integrated bioprocesses can operate optimally while minimizing cost, respecting product quality and productivity constraints, and being robust to uncertainties in an unknown environment.”

Learning and predicting the behavior of bioprocesses in the presence of disturbances from limited and noisy data is of paramount importance for navigating unknown environments, such as Mars, and enabling safe and on-demand biomanufacturing. To this end, the researchers are developing data-efficient approaches by approximating the underlying dynamics of systems via probabilistic models that can be utilized for uncertainty-aware analysis and optimization of safety-critical systems.

George continues, “By leveraging recent advances in computational modeling, machine learning and optimal control, we have created pipelines for process development and optimization on earth and knowledge transfer and adaptation to unknown environments such as Mars. The key impact that stems from this line of work towards human-based exploration of space is the translation of traditional practice in process systems engineering and control to the novel field of space systems bioengineering.”

The research is also focused on the development of methods for the safe, robust, and verifiable composition of machine learning and optimal control in feedback loops for real-time mission uncertainties. Fusing physics-based knowledge of bioprocesses with data-driven models, the research is showing promising results for automated model and controller learning “on-the-fly” using real-time data.

Read more about their research here: cubes.space/resources

*This research is being reported as part of the Berkeley CUBES SDID program (Systems Design and Integration Division) whose mission is to optimally allocate and utilize Mars resources, to tightly integrate and automate internal processes, and to satisfactorily achieve performance per mission specifications.