Serendipitous Stellar Occultations

Global Nominee

Serendipitous Stellar Occultations received a Global Nomination.

THE CHALLENGE: Near Earth Objects Machine Learning

Solar System

For this challenge, we invite you to become "virtual contributors" to the Asteroid Grand Challenge and develop a hypothetical method, concept note or simple prototype that demonstrates how Machine Learning could be used to help us avoid the same fate as the dinosaurs.

Explanation

Near Earth Objects (NEOs) are exceedingly difficult to find through conventional astronomy. With their small size and rapid movement, only the bigger ones (more than 1KM in diameter) have reliably been found.

To date, asteroid detection projects (NEAT, LONEOS, CINEOS, etc.) have focused on using very big and expensive telescopes, which still struggle to find smaller NEOs. Very few have ever been found, yet it’s estimated there are hundreds of thousands of them.

We propose a change of paradigm. By using thousands of low cost telescopes set out across a large area, we will watch for occultations. These are the tiny shadows that asteroids cast as they travel between us and distant stars. From Earth, these are observed as a star that momentarily dims. By using machine learning techniques we can combine the observations from all of the telescopes in real time, search for this pattern, and identify the approximate speed, direction and size for these ‘serendipitous’ events.

Doing this we transfer the detection problem away from expensive telescopes, onto cheap computer resources. And, by using low cost telescopes we can engage the amateur astronomers in the community, who already have the knowledge, enthusiasm and equipment to contribute. We just need to help them get set up to take the right observations and coordinate them through the internet.

For the Space Apps Challenge, we’ve taken the first steps by developing:

A conceptual solution and modelling
A simulated set of observations, including noise
A software solution to search through the simulated observations and find our asteroids

And now can: