By Yan Fisher, Global Evangelist, Emerging Technologies, Red Hat
Undoubtedly, landing on the Moon 50 years ago has to be one of humanity’s greatest technological achievements. The much-written about technologies that took us to the Moon pale in comparison to the phones we have in our pockets today. But, what were the technology innovations that got us to the Moon, and what are the innovations that will take us to Mars?
The computer that got us to the Moon was the size of a briefcase, and there had never been anything like it before - a portable digital general-purpose computer. It was this miniaturisation of computing components, thanks to an electrical engineer from Texas Instruments, Jack Kilby, who invented the integrated circuit, that enabled computers to be installed on spaceships and the lunar modules.
The Apollo 11 mission was the first time software ran on the Moon. Ultimately, it was this combined work of NASA and the Massachusetts Institute of Technology (MIT) that helped give rise to the digital age in space.
Preparing for the next phase of space exploration
It is probably not coincidental that HPE chose their own Apollo servers to build the Spaceborne Computer, bringing one teraflop of computational power from Earth to space. The Spaceborne Computer is an off-the-shelf system installed in a special enclosure so it can practically fit on theInternational Space Station (ISS). Beyond that, Spaceborne runs non-hardened hardware and software, and is controlled by an open source operating system - to be specific, Red Hat Enterprise Linux.
Computing systems on board spacecraft are, usually, highly-specialised and specifically hardened to protect against exposure to cosmic rays, gravitational forces and other environmental hazards. However, ever since the first manned spaceflight in 1961, opinions started to form about the hardening and protection of the hardware. If humans can sustain severe environmental changes for extended periods of time then computer hardware should be able to as well.
HPE and NASA originally planned Spaceborne’s mission to be a year-long experiment, about the amount of time it would take a spacecraft to reach Mars. The goal was to run compute and data-intensive applications in a changing radiation climate and determine the effects of solar radiation on the systems while running. On June 4, 2019, after spending 615 days on board ISS and having traveled nearly 228 million miles, the Spaceborne computer was successify returned to Earth by SpaceX’s Dragon 9 spacecraft.
As we look to reaching the next milestone of space exploration, such as colonising Mars, the outcomes of this project will help scientists to find new ways of using off-the-shelf hardware in space without the need for expensive and bulky protective shielding. It also confirms that commodity computers using standard operating system and software can be used to transport humans to Mars. These machines could then be delivered to the surface of the Red Planet and deployed by scientists and ground personnel to conduct research and experiments.
Increasing openness and collaboration
Back on earth, research institutions and national labs across the globe are pouring hundreds of thousands of research hours into every conceivable aspect of space science. And, overwhelmingly, the high performance computing (HPC) systems used for research are running open source software.
In fact, 100% of the current TOP500 supercomputers run on some form of Linux, with top two, Summit and Sierra, running Red Hat Enterprise Linux. It really isn’t out of the realm of possibility that the ecosystem of space exploration could, therefore, be built on the foundations of sharing the information and contributing to the common knowledge source, much the same as an open source software ecosystem works.
The key elements of success for that ecosystem are:
- The use of open technologies to stimulate collaboration among nations
- AI-driven scientific research
- Active participation from commercial companies, like SpaceX, that would augment the national efforts and provide additional funding for academic research.
While these are tall goals that could take multiple years to achieve, we are already seeing great strides made in every aspect, open source software is already running in space, AI/ML is used in spacecraft communications and navigation, and the number of commercial companies interested in space economy is growing.
Extending open source innovation to hardware design
We could speculate that computer hardware could follow the same pattern as software and a joint effort by multiple parties using an open source design principles, like those used in RISC-V, will create processors to run the brains of a spacecraft or landing module.
Lowering the barrier of entry for electronic design is one of the main goals of an initiative launched by DARPA, which is aiming to work with the microelectronics community to forge a collaborative, cost-shared research agenda to usher microsystems into a new age of innovation. DARPA is helping, to a certain extent, open-source hardware designs.
The DARPA project could bring the design point of a processor from billions of dollars in research and development costs, to tens of thousands, or even less; from the 2-3 years it takes a new piece of silicon hardware to be ready for production, to delivering in a matter of weeks.
Look at how much and how fast the world is changing - it’s about creating a standard set of tools for hardware design that are applicable and accessible. As DARPA itself states, “we need to break away from tradition and embrace the kinds of innovations.”
Adapting to new computational demands
Just like commoditisation and standardisation transformed supercomputers from proprietary to more open designs, if we are to truly make the trip to Mars and manned space exploration possible, we will need a fundamental shift in the way we approach computing infrastructure.
When you think about some of the computer systems that have been on board the ISS, they are 20-25 years old. Once they are there, in space, they usually stay there. Computers we use on Earth today are thousands of times more powerful than computers that run in space.
This is where, for extended space missions, the idea of composable infrastructure becomes very interesting. Composable infrastructure treats compute, storage and network devices as pools of resources that can be provisioned as needed and in real time, depending on what different workloads require.
The approach is not unlike a public cloud in that resource capacity is requested and provisioned from a shared pool – except composable infrastructure sits on-premises in an enterprise data center. Or, in this case, on board the spacecraft. As you take a spacecraft from earth to orbit, and then to distant planets, the purpose and computational needs change progressively.
For instance, in a colonisation effort such as Mars One, once the module lands, it’s not going to leave Mars. Onboard computing systems, therefore, need to be a portable cloud, if you will, that can be self-aware and able to intelligently reconfigure itself down to the basic elements like CPUs, memory and storage, and run general-purpose operating system and orchestration software.
Extending human capabilities with AI
The most valuable currency in the world is humans’ time (man hours), especially in solving problems. Machine learning and artificial intelligence, in general, are transforming industries by allowing humans to spend their time focusing on high-value problems.
In space, these technologies could truly be transformative, as computers could collect, analyse and act upon data acquired during flight without having to involve a human. It is incredibly expensive to send someone into space - NASA announced in June that it would open the International Space Station to private individuals and commercial business at roughly $35,000 per night per astronaut.
If you can take away the need for a computer technician or engineer, or one of the many other hats an astronaut needs to wear, through composable infrastructure and AI, you could make room for more specialists instead of highly-skilled jack-of-all-trades. So from a mission perspective, it would mean that you can send more explorers and scientists with essential skills to colonise Mars, for example.
Embracing innovation and breaking away from tradition is what helped humanity land on the Moon. Getting to the Red Planet and beyond will require a fundamental shift in the way we deploy off-the-shelf, modular and self-learning computer infrastructure; in the way we design hardware; and an entire ecosystem of companies working together to push the very limits of what is thought possible. It was this spirit, this willingness to push the boundaries, that made the Apollo missions a success.
We need to harness that same spirit once more and there is no better place to find innovation and collaboration than in open source communities. When we are advancing the very boundaries of humanity, openness and transparency are key to success and it is why NASA itself is solving some of its greatest problems with open software and crowdsourcing.