Dust Storm: The cost-saving benefits of a compute cloud on Mars

• Published in: Acta astronautica Vol. 168; pp. 31 - 36
• Main Authors: March, Benjamin H., Casler, James G.
• Format: Journal Article
• Language: English
• Published: Elmsford Elsevier Ltd 01.03.2020; Elsevier BV
• Subjects: Cloud computing; Cost control; Data storage; Data transfer (computers); Deep space; Dust; Dust storms; Economics; Mars; Mars dust; Mars missions; Multiagent systems; Planetary industrial operations; Pricing; Sensitivity analysis; Storms; Technology; Economics; Cloud computing; TCO; FSO; Deep space; SSD; PaaS; AI; RadHard; Planetary industrial operations; Mars; COTS; NPV; AWS; GNC; SP; DSN
• ISSN: 0094-5765; 1879-2030
• DOI: 10.1016/j.actaastro.2019.11.029

As more missions to Mars require greater autonomy of both rovers and humans, more sophisticated computer technology in close proximity will be necessary to sustain operations. On Earth, cloud technology has provided significant cost-saving benefits to companies that require heavy compute resources due to the pay-per-use pricing model. The same could be true on Mars, assuming that there is enough demand for high-capacity, on-demand compute and data storage resources. The concept of a compute cloud about Mars (“Dust Storm”) was developed to determine a basic pricing model for compute, and other Platform as a Service (PaaS) features. Existing mission concepts were re-evaluated under the assumption of using Dust Storm versus developing and deploying an in-house solution. Additionally, a sensitivity analysis was conducted to control for inaccurate assumptions and estimates. It was determined that five-year multi-agent missions would cost significantly less (48.6%) when renting time from Dust Storm. It was also determined that the greatest effects on cost savings are the cost of building and deploying Dust Storm, the maximum data transfer rate between Earth and Mars, and the length of time expected to recoup initial investments. Future research will be required to reduce hardware costs, and increase data transfer rates. •The number of multi-agent missions to Mars will increase in the coming decades.•A tenant-based cloud infrastructure at Mars could be used to coordinate missions.•Developing and managing a cloud infrastructure at Mars would be a profitable business.•A combination of COTS and RadHard hardware will reduce costs and increase performance.•Additional research into high-performance space computers is still needed.