April 5, 2018 : Scientists Resort To Artificial Intelligence To Assess The Potential For Life On Alien Worlds Like Titan
A group of researchers from the University of Plymouth reveals the potential interest of artificial intelligence in our search for extraterrestrial life on alien worlds in the Solar System and around other stars. Artificial intelligence that is rapidly developing in various sectors can also help us identify exoplanets which have the ingredients or the environmental conditions to host life, from typical lifeforms to more exotic lifeforms. The project mobilizing artificial intelligence to find potential biospheres, here, is proposed by the PhD student Christopher Bishop and his supervisor Angelo Cangelosi who is a Professor in Artificial Intelligence and Cognition. Both specialists are developing an artificial neural network (ANN) based on multiple parameters linked to potential habitability. Their system allows them to classify planetary bodies into five types. Once the researchers have identified the type of world, they are in a position to evaluate the probability of life.
The artificial neural network developed by Christopher Bishop and Angelo Cangelosi, who are based in the Centre for Robotics and Neural Systems, can clearly play a major role in future missions towards other stars like Proxima Centauri or Sirius. We know, now, that there is a planetary body evolving in the Habitable Zone of Proxima Centauri. Are there pools or oceans of liquid water on the surface of that exoplanet known as Proxima Centauri b ? Christopher Bishop, who is also a Lecturer in Radio and Telecomms at the Britannia Royal Naval College in Dartmouth, had to describe his work at the European Week of Astronomy and Space Science (EWASS) in Liverpool. The artificial neural network developed by Christopher Bishop and his collaborator has been trained to classify planetary bodies on the basis of their type or category. The different types of planetary bodies identified by both specialists are the type of the present-day Earth, the type of the early Earth, the type of Mars, the type of Venus and the type of Saturn's largest moon Titan.
The five types of worlds or planetary bodies are all rocky bodies and are covered by an atmosphere. The atmosphere of the early Earth was probably very different from the atmosphere of the present-day Earth. The oxygen present in our atmosphere is closely related to our biosphere which engenders oxygen via photosynthesis in particular. The atmosphere of Mars, dominated by carbon dioxide (CO2), is particularly thin with a relatively low atmospheric pressure on the surface. The atmosphere of Venus, dominated by carbon dioxide, is particularly thick and dense with a relatively high atmospheric pressure on the surface. The atmosphere of the Opaque Moon Titan, dominated by molecular nitrogen and containing a relatively high concentration of methane in the Troposphere, is relatively deep and thick with an atmospheric pressure on the surface which is higher than that of the Earth at sea level. The composition of Titan's surface may be quite different from that of Earth's surface since the environment of Titan is rich in water ice and hydrocarbons. The five types of planetary bodies, identified by both researchers and located in our Solar System, may be among the most interesting worlds in terms of exobiology since they are potentially habitable.
What are artificial neural networks concretely ? Artificial neural networks represent systems that try to replicate the way the human brain learns. They can identify complex features or patterns that are too complex to analyze for our brain. Artificial neural networks represent one of the main tools mobilized in machine learning that is the field of computer science that uses statistical techniques to bring computer systems or robots the ability to learn and to make the right decision. The system developed by Christopher Bishop and Angelo Cangelosi is likely to form the basis of a robotic probe which would have the capacity to prioritize its actions or its goals in terms of exploration of new planetary systems or exoplanets. The probe would also have the potential to identify new biospheres or potential biospheres. In fact, the probe would have, to a certain extent, its own intelligence and would be autonomous. That's very important for remote places located several light years away since the instructions from our planet can't be sent instantaneously. For instance, the probe would have to avoid any debris disk around any planet or Gas Giant.
Christopher Bishop pointed out : « At the moment, this project is all about investigating the feasibility of the idea. But we believe artificial intelligence would play a key role as such a probe would be out of range of any effective human contact and would need to make decisions on its own. We're also looking at the use of large area, deployable, planar Fresnel antennae to get data back to Earth from an interstellar probe at large distances. This would be needed if the technology is used in robotic spacecraft in the future. » The closest star to our Solar System is Proxima Centauri and it is located about 4 light years away from us. That's in fact a huge distance and any interstellar probe would have to be much faster than the conventional planetary probes we know today. Researchers are currently studying the possibility of propelling tiny probes on the basis of powerful lasers. That study is part of the project Breakthrough Starshot. If we manage to send a probe which can be extremely fast, we'll also have to find a solution to slow down the probe once it reaches the area of the planetary target.
Christopher Bishop has already put his skills at the service of NASA's Goddard Spacecraft Centre to generate simulated data based on observations captured by observatories located on Earth and by orbiting observatories. The spectrum of each atmosphere of the five Solar System worlds, selected by both specialists of artificial intelligence, represents an input to the artificial neural network. The system is then in a position to classify the studied worlds in terms of the planetary type. How can we determine that a world is likely to harbor life since life has only been found on Earth until now ? In fact, the classification proposed by both researchers is based on a « probability of life » mark. The mark or metric has relatively strong foundations in our understanding of atmospheric and orbital properties of the five key planetary worlds of our Solar System, which can harbor life, which may have harbored life or which may become habitable in the future.
Christopher Bishop has trained his system or ANN with over 100 different spectral profiles and each profile incorporates several hundred parameters that are likely to be in line with the potential of habitability. So far, the system of artificial intelligence works well or leads to a satisfying outcome when presented with a test spectral profile that it hasn't encountered before. The Professor Angelo Cangelosi, who has performed an extensive work comprising some studies with the European Space Agency and NASA, argued : « Given the results so far, this method may prove to be extremely useful for categorising different types of exoplanets using results from ground-based and near Earth observatories. » Future space observatories like ESA's Ariel Space Mission and NASA's James Webb Space Telescope may allow us to obtain spectral data of higher resolution which means that we will be in a better position to characterize exoplanets, in the near future. The artificial neural network proposed by both collaborators may be useful to select targets in the near future.
In our Solar System, a planet like Venus is located in the Habitable Zone and the planetary body looks like the Earth in terms of size, mass and mean density. However, the opaque and dense atmosphere of Venus generates a huge greenhouse effect which prevents the presence of liquid water on the surface. Surface temperatures on Venus are around 460 degrees Celsius and the atmospheric pressure on the ground of our sister planet is around 92 Bar. Was Venus habitable in the past ? Will it become habitable one day ? Why is Venus so different from the Earth ? The hypothesis of the presence of ecosystems on Venus has not been ruled out. The Venusian atmosphere may contain colonies of bacteria or extremophiles according to certain exobiologists. From outer space, the surface of Venus is completely hidden by its opaque atmosphere which is almost uniform in the visible spectrum.
In fact, from outer space, the appearance of Venus is closer to the appearance of the Orange Moon Titan than to the appearance of the Earth. Yet, Titan is very different from Venus since the atmosphere of Titan is mainly composed of molecular nitrogen like the atmosphere of our planet whereas the atmosphere of Venus is dominated by carbon dioxide. The clouds of Venus are composed of sulfuric acid whereas the clouds of Titan are dominated by methane or ethane. Some planetologists have advanced the possibility of a methane-based life on Titan since there is a methane cycle comparable to the water cycle of the Earth. Radar data acquired from the Cassini probe during its long mission have clearly shown the presence of seas, lakes or rivers of methane or ethane in the high latitudes of Titan. Titan is rich in organics and hydrocarbons. Complex molecules can take shape in the environment of Titan. The haze of Saturn's largest moon is likely to teach us a lot regarding the prebiotic chemistry. That's why the worlds which look like Titan must be considered exobiologically speaking.
This composite view, generated on the basis of infrared data acquired with the Visual and Infrared Mapping Spectrometer of the Cassini orbiter during the T-114 flyby or the flyby of November 13, 2015, reveals surface features on Saturn's largest moon Titan. One can notice in particular the dark regions Fensal and Aztlan which mark a sharp contrast with the surrounding area whose mean albedo is higher. The Opaque Moon may have the highest habitability rating among known planetary bodies beyond the Earth, on the basis of parameters such as availability of energy and several surface and atmospheric characteristics. The artificial neural network developed by Christopher Bishop and his collaborator may help us better identify potential biospheres or key candidates for habitability. Image Credit: NASA/JPL/University of Arizona/University of Idaho.
- To get further information on that news, go to: https://www.plymouth.ac.uk/news/researchers-use-artificial-intelligence-to-predict-probability-of-life-on-other-planets and http://www.spacedaily.com/reports/Artificial_intelligence_helps_to_predict_likelihood_of_life_on_other_worlds_999.html.