AI could help the Europa Clipper mission to make all kinds of discoveries!

This presents many challenges, many of which stem from the fact that the Europa Clipper will be very far from Earth when it carries out its scientific operations. To solve this problem, a team of researchers from the Jet Propulsion Laboratory (JPL) of NASA and the University of Arizona (ASU) has designed a series of machine learning algorithms that will enable the mission to to explore Europa autonomously.

The way these algorithms could facilitate future deep space exploration missions was the subject of a presentation last week (August 7) ​​at the 25th ACM SIGKDD conference on knowledge discovery and data mining in Anchorage, Alaska. This annual conference brings together researchers and practitioners of science, exploration and data analysis from around the world, to discuss the latest developments and applications.

Basically, communicating with missions in deep space is a tedious and difficult job. When you communicate with missions on the surface of Mars or in orbit, the signal can take up to 25 minutes to reach them from Earth (or vice versa). On the other hand, sending signals to Jupiter can take between 30 minutes and one hour, depending on its location in its orbit relative to the Earth.

As the authors note in their study, the activities of spacecraft are usually transmitted in a predefined script rather than through real-time commands. This approach is very effective when the position, the environment and other factors affecting the spacecraft are known or can be predicted in advance. However, it also means that mission controllers can not react in real time to unexpected developments.

As Dr. Kiri L. Wagstaff, Principal Investigator in NASA's JPL Machine Learning and Instrument Autonomy Group, told Universe Today by e-mail:

"To explore a world too far away to allow direct human control is a challenge. All activities must be pre-scripted. A quick response to new discoveries or changes in the environment requires that the spacecraft itself make decisions, called spacecraft autonomy. In addition, operating at one billion kilometers from Earth, data transmission rates are very low.

"The ability of the spacecraft to collect data exceeds what can be returned. This raises the question of what data should be collected and how to prioritize it. Finally, in the case of Europa, the spacecraft will also be bombarded with intense radiation, which can corrupt the data and cause computer resets. Coping with these dangers also requires autonomous decision-making. "

For this reason, Ms. Wagstaff and her colleagues began to research possible methods for analysis of on-board data that could be used at any time and at any time when direct human monitoring is impossible. These methods are particularly important when there are rare and transient events whose occurrence, location and duration can not be predicted.

These include phenomena such as the dust devils that were observed on Mars, the impacts of meteorites, the lightning on Saturn and the iced plumes emitted by Enceladus and other bodies. To solve this problem, Ms. Wagstaff and her team looked at recent advances in machine learning algorithms, which enable a degree of automation and independent decision making in computer science. As Dr. Wagstaff said:

"The automatic learning methods allow the spacecraft itself to examine the data as and when they are collected." The spacecraft can then identify which observations contain events of interest. This can affect the allocation of downlink priorities. The goal is to increase the chances that the most interesting discoveries will be read first. When data collection exceeds what can be transmitted, the spacecraft itself can tap the extra data to find valuable scientific nuggets.

"On-board analysis can also allow the spacecraft to decide what data to collect later based on what it has already discovered. This has been demonstrated in Earth orbit with the help of the Sciencecraft standalone experiment and on the surface of Mars using the AEGIS system on the Mars Science Laboratory rover (Curiosity). Autonomous and responsive data collection can significantly accelerate scientific exploration. Our goal is also to extend this capacity to the external solar system. "

These algorithms have been specially designed to facilitate three types of scientific investigations of extreme importance for the Europa Clipper mission. These include the detection of thermal anomalies (hot spots), compositional anomalies (minerals or unusual surface deposits) and active plumes of icy matter from the oceanic subsoil Europa.

"In this context, the calculation is very limited," said Dr. Wagstaff. "The spacecraft computer runs at a speed similar to that of a desktop computer in the mid to late 1990s (about 200 MHz). We therefore favored simple and effective algorithms. An added benefit is that the algorithms are easy to understand, implement, and interpret. "

To test their method, the team used their algorithms on both simulated data and observations of space missions to the moons of Jupiter and other planets in the solar system. These included the Galilee spacecraft, which made spectral observations of Europa to determine its composition; the Mars Odyssey spacecraft, which searched for thermal anomalies on Mars; and the The Hubble Space TelescopeObservations of the activity of the plume on Europa.

The results of these tests showed that each of the three algorithms had a sufficiently high performance to meet the scientific objectives defined in the global survey on the decennial science of 2011. These include "confirm the presence of a inner ocean, to characterize the satellite ice shell and understand its geological history 'on Europa in order to confirm' the potential of the outer solar system as a residence for life '.

In addition, these algorithms could have considerable implications for other robotic missions to distant destinations. Beyond Europa and the moon system of Jupiter, NASA hopes to explore the moons of Saturn Enceladus and Titan in search of future signs of life, as well as even more distant destinations (such as Neptune's moon Triton and even Pluto ). But applications do not stop there. As Dr. Wagstaff says:

"The autonomy of spacecraft allows us to explore where humans can not go. This includes remote destinations such as Jupiter and locations beyond our own solar system. It also includes closer environments that are dangerous to humans, such as the bottom of the ocean floor or the high radiation parameters on Earth. "

It is not hard to imagine a near future where semi-autonomous robotic missions will be able to explore the outer and inner reaches of the solar system without regular human surveillance. Further into the future, it's not hard to imagine a time when fully autonomous robots are able to explore extra-solar planets and send their discoveries home.

And meanwhile, a semi-autonomous Europa Clipper could find proof that we all wait! These are biosignatures that prove that there really is a life beyond the Earth!

Suggested Readings: KDD 2019 Study (PDF)