“I know I’ve made some very poor decisions recently, but I can give you my complete assurance that my work will be back to normal. I’ve still got the greatest enthusiasm and confidence in the mission. And I want to help you.” - Hal 9000 (to Dave Bowman, mission commander)

It’s been over 50 years since Hal, the supercomputer powering a spacecraft in Stanley Kubrick’s “2001, A Space Odyssey” uttered those words. Fast forward 50+ years and we are beginning to experience what Kubrick foresaw so many years ago in Hal. Over the last few years cheaper launch technology and miniaturization of satellites have enabled the commercialization of space as well as open the path to space computing, the latest dimension of edge computing that brings computation and data storage closer to the sources of data. Edge computing basically moves the processing and analyzing capabilities away from the data center and closer to the actual data source…in the case of space computing – to the satellite.

Commercial imaging and Earth observation (EO) were among the first applications to realize the potential of space (or onboard) computing. The image data captured by these satellites is used in multiple applications, including mineral deposit extraction, disaster mitigation and recovery, agriculture development and management and intelligence gathering, among others. Traditionally, satellites’ role has been to collect the data and then transmit it back to Earth to extract the usable data and delete the rest. Onboard computing streamlines this process as vast volumes of imaging data no longer need to be downloaded…the analysis and extraction are done in the satellite. The efficiency of space computing saves both time and bandwidth as data storage and downlink transmission are managed more efficiently when fewer images and less data need to be transmitted.

But space computing is not without its challenges, chief of which is getting there. Access to space has become more affordable as launch technologies and methods including routine use of reusable launch vehicles, increased rideshare opportunities and newer small vehicles (operational and planned) have opened satellite launches to a wider segment. Once launched, the next challenge is staying there. Data centers consume much energy and new methods of providing a continuous flow of energy to satellite space computing centers will need to be developed as will the ability to maintain exact control of temperature as both will be critical to the health and well-being of space data centers.

But despite the challenges, companies ranging from IBM and Hewlett Packard Enterprise to startups around the world are vying to position themselves to capture a piece of the nascent, but rapidly growing space computing industry, which could evolve from onboard satellite computing to orbiting data centers in space.

Spiral Blue, an Australian company, processes images on the satellite as they are collected using a mixture of old school remote sensing techniques and modern artificial intelligence (AI) techniques to process raw image data into information, reducing the size of images. This can result in a data reduction of 20-1000 times while retaining all the desired information of the original raw image. This helps overcome the bandwidth problem, creating benefits in cost, flexibility, and lead time for services derived from satellite images. According to Taofiq Huq, the CEO of Spiral Blue, today’s EO satellites are quite inefficient and have many limitations, chief among them is that they are unable to send down as much data as they collect due to limitations in bandwidth. Satellites are sending large volumes of raw data down to Earth for processing, says Huq. This means there is a lot of wasted capacity of a satellite. With edge computers, satellites will become more efficient by bringing down the cost of imagery, providing insights in a timelier fashion, and giving end-users more customization as to how they access and utilize remote sensing data.¹

Hewlett Packard Enterprise’s Spaceborne Computer-1 answered three crucial questions regarding the future of space computing:

• can you take off-the-shelf components computing components and put them into space?
• can they survive the launch and be installed by non-IT experts?
• can they function in space – and if so, for how long?

When SBC-1 returned to earth after almost two years on the International Space Station, the answers were Yes, Yes and Yes…and SBC-2 was launched last year. During its mission, SBC-2 will assist ISS National Lab users in conducting experiments that demonstrate the computer’s proficiency, and further demonstrate onboard state-of-the-art artificial intelligence (AI) and edge computing potential.²

Switzerland’s RUAG Space is teaming up with a software provider to run artificial intelligence solutions on its Lynx computer, which it says is the most powerful commercially available onboard satellite computer. According to Anders Linder, head of RUAG Space’s global satellite business RUAG, the company’s Lynx computer is “250 times more powerful than normal on-board computers,” including those the company currently delivers for European Space Agency programs. Additionally, according to Linder, giving satellites more computing power also enables more powerful artificial intelligence solutions.³

The Space Development Agency, (SDA), which supports space development in the interests of U.S. national security is taking a different approach. “On the ground, I can tie myself to a hydroelectric dam and a river to cool my processing center, but in space, you’re always going to be limited by how much heat you can dump and power you can collect,” said Derek Tournear, who leads the Space Development Agency, at a recent webinar. To assemble enough computing power to do machine learning in space, you need to put a lot of small computers in low Earth orbit and then link them up. Over the next two years, a DARPA program called Blackjack will attempt to prove out concepts that could be used to build a self-organizing orbital mesh computing network.

Looking to the future, in-space computing will go far beyond imaging and EO, but also be critical to the success of missions to Mars and beyond. “If we can take computational resources with us on our mission and they can give us the correct answers, then we can collect the data, process it, and hopefully come up with the necessary insights to continue our mission. New computer chip architectures, perhaps mimicking the human brain, could accelerate the creation of an orbiting machine learning network, said Jeff Sheehey, chief engineer of NASA’s Space Technology Mission Directorate. His agency is building processors that are “100 times better” than the radiation hardened ones that power today’s spacecraft. “We’re also looking at neuromorphic processors,” he said, chips that function less like conventional integrated circuits and more like the synapses of the human brain. More intelligent systems will be necessary if humanity is to build sustained presence in space, he said. It’s “going to require a lot of assets to be in place that may not have humans tending them for long periods of time. The humans will be there intermittently.”

Could we one day hear another space computer utter one of Hal’s famous lines revealing that it was not the helpful computer we first heard from:

“I’m sorry Dave, but in accordance with sub-routine C1532/4, quote, ‘When the crew are dead or incapacitated, the computer must assume control’, unquote. I must, therefore, override your authority now since you are not in any condition to intelligently exercise it.”