With too many jobs, not enough gravity, and a cramped living space offering no prospect of going outside for a quick jog, life as an astronaut on the International Space Station can be hard work. Could a flying robot buddy help take some of the load? That’s what IBM, Airbus, and Germany’s DLR space agency, the creators of a newly deployed robot called CIMON, are hoping.
Delivered to the International Space Station this week aboard December 5’s Falcon 9 resupply mission, the new, improved CIMON robot promises to help out with astronauts’ daily experiments in the European Columbus research module of the ISS — and maybe even offer a bit of moral support as well.
CIMON-2, as it is technically known, is the successor to the original CIMON robot deployed on the ISS beginning in 2018. CIMON version 1 returned to Earth in August 2019 to make way for its successor. According to official documentation, CIMON is an acronym standing, somewhat tenuously, for Crew Interactive Mobile companiON. Depending on who is telling the story, it may also be a reference to the character of Professor Simon Wright from the pulp novel and anime series Captain Future.
Like that character — the talking, living brain of a scientist inside a floating, transparent case — CIMON is, in essence, a flying brain. It’s a spherical, free-flying robot, powered by artificial intelligence, which uses a touchscreen and voice commands to communicate with astronauts. Picture one of the screen-equipped Amazon Echo Show devices floating through the ISS and you won’t be too far off!
“The base technology is very similar,” Matthias Biniok, who worked as an IBM project lead on the CIMON-2 robot, told Digital Trends. “We use our IBM Watson intelligence to do similar things to what Siri, Google Assistant, or Alexa are doing: understanding human language and reacting based on what the human is saying. The cool thing about CIMON is that he is trained specifically for astronauts. He understands specific language, based on the things he will be asked to do on the ISS. He also knows about procedures and documentation, and can provide systems and information to support this.”
The ISS can be surprisingly loud at times, which doesn’t always make things easy for a voice-activated A.I. assistant. An “acceptable” level of noise during the ISS’s 16-hour work day is 72 dB, slightly louder than a continuously running vacuum cleaner. Some former astronauts have even reported hearing loss after stays on the station. To help make sense of speech with that kind of background noise, CIMON-2 uses an array of eight microphones with which to direction sound has come from. It also has an extra directional microphone to improve voice comprehension. For playback, its “mouth” is a loudspeaker which can be used for speaking or playing music.
CIMON-2 can do a lot more than that, though. After all, why build a physical robot if all you’re going to be doing is receiving voice queries and offering verbal responses? In that case, it would be much more straightforward to incorporate additional microphones and speakers into the existing ISS equipment; especially since this would cut down on the number of free-floating objects in space. However, having an actual robot opens up all kinds of possibilities. CIMON-2 can indicate responses by either nodding or shaking its head, eschewing the need to always communicate via speech.
It can additionally carry out tasks like moving about the ISS searching for objects on demand, or taking stock of inventory. To move, it uses 12 internal fans, which let it travel freely in all directions in the ISS’s microgravity environment. It can move autonomously, on command, or by following the path of astronauts as they go about their daily activities. To avoid potential collisions, it employs on-board ultrasound sensors.
The robot assistant serves a valuable role when it comes to documentation. CIMON-2 features twin cameras capable of identifying objects or crew members using facial recognition. It boasts five additional cameras for orientation and recording the world around it. Should an astronaut need, for instance, to get an image of something happening in a different part of the ISS, they can ask CIMON-2 to take it and then return.
“The batteries are good for two hours,” Biniok said. “Afterwards, they need to be recharged. But there are several battery packs up there, so you could have CIMON run for a long time.” In the future, he said, the team is considering a system that will allow CIMON to autonomously return to a charging station when it is running low on juice.
Another reason for giving CIMON-2 a physical body is because of a use case that doesn’t actually exist yet. CIMON-2’s creators note how, for the first time, their robot has emotion recognition capabilities: scouring the content of sentences and the tone for clues about the astronauts’ emotional state.
This could lead to it being more useful by changing its behavior based on the response of the humans around it. Going forward, it could have more profound applications as well. “In the future it could be that CIMON [may also be used as a friend to those on board],” Biniok said. “This may have positive psychological effects.” While astronauts are selected and trained to be extremely mentally tough, it’s still a challenging environment in which to be stuck for long periods of time in a confined space with just a few other people. This effect could intensify in the future as astronauts are potentially used for long haul space missions.
Building a robot buddy might sound like something from a Pixar movie, but it could be a valuable tool for astronauts. “It’s very important that we have something like CIMON that could work with the astronauts to support them on an emotional and psychological level,” continued Biniok. Unlike a disembodied voice, CIMON’s physical form gives astronauts something more tangible to interact with.
It is, of course, impossible to talk about spaceship-based A.I. assistants without mentioning HAL 9000. The main antagonist in 2001: A Space Odyssey, HAL is the calm, conversational artificial intelligence whose job it is to assist the crew and maintain the ship’s functions; only to turn on its crew and begin trying to kill them off. Anyone who builds a smart assistant for astronauts is aware of the (somewhat troubling) pop culture legacy of doing such a thing. In the case of IBM, however, it is even more notable.
Although HAL is ostensibly an acronym derived from “Heuristically programmed ALgorithmic computer,” it didn’t take long after the 1968 movie’s release for some fans to realize that HAL is a one-letter shift from IBM. Writer Arthur C. Clarke said that he and Stanley Kubrick did not realize this at the time. “As it happened, IBM had given us a good deal of help, so we were quite embarrassed,” he later noted. “[We] would have changed the name had we spotted the coincidence.” Nonetheless, the association has stuck; something that Biniok acknowledged when designing CIMON.
“Of course, you can ask CIMON to ‘open the pod bay doors’ and he will answer, ‘I’m afraid I cannot do that,’” he laughed. “There are some jokes in there — and not just from 2001: A Space Odyssey, either.”
On a more serious note, he said that the team was engaged with the central question of how to develop a robot assistant that helps, rather than hinders, in space. “It is something that we kept in mind when building CIMON,” Biniok said. “For example, on the back of CIMON there is an offline button and a power button. [We did our best to consider the] ethical aspects. The astronauts need that as a reminder that they are always in charge. CIMON is always just an assistant.”
For now, CIMON exists as a “technology demonstrator.” While he helps with experiments in space, he also stands (or, well, floats) as something of an experiment himself. There have been mishaps along the way, such as a video in which the bot appeared to get testy with an ISS astronaut. Asked to stop playing the song Kraftwerk’s “The Man-Machine,” version one of CIMON refused and responded that, “I love music you can dance to.” When it was asked again, CIMON responded, “Be nice please,” and, hurt, asked “Don’t you like it here with me?” Not the biggest of glitches, perhaps, but perhaps a little concerning given its familial connection to HAL.
“[The first iteration of CIMON] was very successful, but as with all missions there are things that we learn that we want to improve,” Biniok noted. “This time we’ve implemented both hardware and software upgrades, which are based on learning from the first version.”
Overall, though, he said that CIMON has been received incredibly. Well. “[The astronauts on the ISS] actually said that they would like to work with CIMON again,” he said proudly. “I think that’s one of the best compliments you can get if you’re working on such a project.”
- Analog A.I.? It sounds crazy, but it might be the future
- The funny formula: Why machine-generated humor is the holy grail of A.I.
- Here’s what a trend-analyzing A.I. thinks will be the next big thing in tech
- Read the eerily beautiful ‘synthetic scripture’ of an A.I. that thinks it’s God
- The future of A.I.: 4 big things to watch for in the next few years