Nowadays, artificial intelligence plays a central role in the technological world, but also more generally in our lives. Developments in recent years ensure that complex problems can be solved in unimaginable timescales. To give a few examples, we can cite the prediction of the particular structure of certain proteins using machine learning techniques, a problem that has plagued scientists for more than 50 years. Or the application of sophisticated algorithms to analyse aerodynamic data. In this context, how is artificial intelligence being used in the field of space exploration?
Artificial Intelligence and Space
The exciting advances made by machine learning allow us to make progress in any existing discipline, from engineering to economics. So why not take advantage of this powerful tool outside the walls of planet Earth as well? Let’s take a look at the possible uses and their benefits.
AI as an assistant to astronauts
“Newton’s third law. The only way humans have found to move forward is to leave something behind.” This is the phrase with which we had to say goodbye to Tars during the film Interstellar. The astronauts’ robot assistants Amelia and Cooper (let’s not forget Case) were instrumental in helping the mission succeed.
Clearly, there are no such assistants aboard the ISS in Earth orbit today, although it would be a pleasure for astronauts to spend time with a 100% humorous Tars. However, researchers are working on something similar. They are trying to develop AI-based assistants that can facilitate the daily tasks of crew members.
The first example is Cimon. It is a digital assistant that will be on board the ISS from December 2019. It will be tested for three years and will be called upon to carry out tasks as required by the astronauts. Another digital friend could soon be joining Cimon. In fact, the Johnson Space Center in Houston (NASA) is developing Robonaut2, which will work alongside the astronauts and, despite itself, will have to carry out the most risky tasks of the mission.
Space mission planning
Planning a space mission is certainly something complicated. But, as we know, in the world of space engineering, innovations are not welcome, and people prefer to build on past knowledge and play it safe.
However, where do you get this valuable information from if you have no experience for the mission to be planned? Thinking about the human exploration of Mars, how can you provide the engineering team with all the information they need to produce optimal planning? For these reasons, researchers are working on the creation of a digital assistant. In this case, it would be called upon to store all the relevant information of the space missions of interest, and perform heavy calculations that would take up valuable human time.
Artificial Intelligence Processing of Satellite Data
Earth observation satellites produce a huge amount of data that needs to be properly processed. These data are received by so-called ground stations and have to be collected and analysed by means of appropriate algorithms. In this sense, artificial intelligence can ensure that information can be analysed in a detailed and efficient manner.
The capabilities of AI are also capable of solving data analysis problems of a complicated nature. Examples include the study of heat generation in urban areas, the coupling of meteorological data with satellite imagery to provide estimates of wind properties, or the techniques used to build health monitoring systems for satellites themselves.
The issue of space debris
The issue of space debris characterises one of the most important space challenges of our time. According to the European Space Agency, there are some 34,000 objects of a size and location that could cause serious damage to space infrastructure. Therefore, in recent years, various strategies are being studied to minimise the risks posed by such debris.
Some approaches involve, for example, returning satellites to low orbit and allowing them to disintegrate in the atmosphere. Or use machine learning logic to develop methods of manoeuvre design to avoid collisions of any kind. By doing so, collisions would be minimised or eliminated, and therefore new debris produced. We could think about using the great computational power we have on the surface to develop complex models, and then send them into orbit to be implemented by spacecraft.