皇家华人

笔奴谤辞苍驳辞

Feature

Controlling traffic in space

17 August 2021
With the recent proliferation of small satellites, space is becoming more crowded all the time especially near Earth.

Space. The very word conjures vastness and emptiness. In fact, space is becoming more crowded all the time, especially near Earth. With the recent proliferation of small satellites, there鈥檚 an increasing need to control space traffic. Fortunately, and of the University of Auckland鈥檚 鈥 are working on solutions.

Armellin is a professor and Pirovano is a postdoctoral research fellow, both in the Department of Mechanical Engineering. They are experts in astrodynamics, the study of the motion of man-made objects in space.

 Armellin first got into the field to understand the motion of asteroids and whether they might be dangerous for the Earth. Work he did with the European Space Agency helped determine that an asteroid due to approach Earth in 2029 wouldn鈥檛 pose a danger 鈥 and neither would its return in 2036. With no immediate need to save humanity from the fate of the dinosaurs, Armellin turned to a much more credible danger: space debris.

Image
Laura Pirovano and Roberto Armellin

The space junk problem

Space is admittedly a lot emptier than a city. But just as there鈥檚 more traffic on some roads than others, there are more man-made objects in some regions of space than others.

Low Earth orbit is the 鈥渄owntown鈥 of space because it鈥檚 close to Earth. This means it鈥檚 cheaper and easier to get satellites and astronauts up there.

The next busiest region is geostationary orbit. This is much further out but at a specific altitude: 35,786 km from Earth. Satellites in this orbit travelling at three kilometres per second spin at the same rate as the Earth, allowing them to stay in an apparently fixed position in the sky. This is useful for telecommunications satellites, because an antenna on Earth can always remain pointed at that satellite.

Satellites鈥 lifetimes can span from a few months to decades. Once their mission is over, guidelines suggest removing them from protected regions within 25 years. However, because until relatively recently there were no regulations about what to do with dead satellites, a lot of them are still orbiting Earth in the busy zones. Also, some spacecraft launches were designed to cast off parts. That鈥檚 not even mentioning the loose springs and screws.

鈥淚f we do not know an object is in a certain orbit, it鈥檚 basically a wandering bullet,鈥 says Pirovano. 鈥淚t may hit a satellite, destroy it and create thousands more pieces of space debris. That鈥檚 why it鈥檚 very important to have an accurate catalogue of objects in space.鈥

Cataloguing space debris

A major part of Pirovano and Armellin鈥檚 work is cataloguing space objects.

鈥淵ou can鈥檛 regulate traffic if you don鈥檛 know where the cars are,鈥 Armellin says.

The difficulty for scientists is that though it may be possible to observe unknown objects, they鈥檙e observed for such a short time that it鈥檚 hard to know where they鈥檙e going or if two observations scattered over time are of the same object.

The catalogue must also be maintained because objects may move from their previously known orbits. For that reason, part of Pirovano鈥檚 research is about detecting satellite manoeuvres.

The catalogue Pirovano and Armellin have been working on with international colleagues now contains about 37,000 objects of ten centimetres and above in size.

鈥淚f you go down to a centimetre, though, there are at least 700,000 objects in space,鈥 says Armellin. 鈥淧roblem is, the catalogue is not complete. Considering we now have a lot of cube satellites of ten cubic centimetres, a bullet of one centimetre would completely destroy it.鈥

Space junk

Predicting collisions

For a satellite to avoid a collision, it has to know one is likely. Several private companies now popping up internationally propose to help governmental efforts, says Armellin.

鈥淪atellite operators receive messages that tell you, 鈥榊our spacecraft is going to have a conjunction on this date, at this distance, with a collision probability of this value.鈥欌

One challenge is the aforementioned problem of knowing what鈥檚 out there. Another is accurately determining the trajectory of space objects.

 

鈥淎s activity in space increases, the number of conjunction messages you receive is going to increase significantly,鈥 says Armellin. 鈥淪ometimes the accuracy of these conjunction messages is low, so people might take the risk of not manoeuvring.

鈥淲hen you鈥檙e driving a car, you might feel you鈥檙e at risk when your car is 30 centimetres from another one. Currently, these systems say a spacecraft is in danger when it鈥檚 a kilometre away from another one. So we have a lot of conjunction messages we need to process for events that are actually not dangerous. If we improve the accuracy, then we reduce false alarms, so space traffic control becomes more practical.鈥

鈥淚f we do not know an object is in a certain orbit, it鈥檚 basically a wandering bullet.鈥

Laura Pirovano

Avoiding collisions

Once collisions can be predicted with a higher degree of accuracy, the next challenge becomes figuring out what to do about them. If the predicted collision is with a piece of space debris, the satellite has to move over. If it鈥檚 with an active satellite, though, the two satellites have to coordinate a response.

The problem is, projected collisions could occur with little notice at any time of day or night, potentially between two operators who aren鈥檛 in contact or who don鈥檛 speak the same language. 

The solution? Autonomous manoeuvring. 

With other colleagues, Armellin and Pirovano are working on using artificial intelligence and machine learning to decide what to do when two objects are due to closely approach each other and to perform any necessary manoeuvres. While Armellin and Pirovano aren鈥檛 machine learning experts, they are contributing their knowledge of space objects and their motion in helping improve autonomous systems.

鈥淚t鈥檚 a challenge because the software always needs to have an answer and that answer needs to be reliable,鈥 says Armellin. 鈥淥n top of that, the answer needs to be optimal, because space is expensive, so you always need to use the minimum amount of propellant in the minimum time.鈥

Cleaning up the mess

These days, satellites need to have an end-of-life plan. For satellites in geostationary orbit, the accepted solution is to move further away from Earth into a graveyard orbit. 

鈥淭he problem is that a graveyard orbit doesn鈥檛 remove the problem,鈥 says Armellin. 鈥淚n the future, these regions of space may become of interest to other missions and then the space junk out there may interfere.鈥

Satellites in low Earth orbit are supposed to have enough fuel at the end of their lives that they can be pulled into Earth鈥檚 atmosphere, where small satellites burn up and larger ones can be manoeuvred to crash into the ocean. 

鈥淗owever, operators may tend not to implement end-of-life manoeuvres because basically, you鈥檙e wasting propellant to kill the spacecraft,鈥 says Armellin. Armellin and some of his colleagues have been working on a solution that exploits satellites鈥 orbit perturbations to reduce the change in velocity necessary to make a spacecraft re-enter the atmosphere. 

鈥淚t鈥檚 cheaper and the satellite can continue to do science while in this phase of slowly spiralling down to the Earth,鈥 says Armellin.

Another University of Auckland colleague, Professor , director of , is working on ways to . Because complex problems tend to require multiple solutions, both groups are contributing to solving the space junk problem. 

鈥淲e鈥檙e space engineers so there鈥檚 room for shaping the future,鈥 says Armellin. 

鈥淭here鈥檚 a lot to be done,鈥 agrees Pirovano. 鈥淭hat means there鈥檚 a lot of opportunity.鈥

Want to connect with the University鈥檚 space research?