Skip to content

Latest commit

 

History

History
213 lines (172 loc) · 10.7 KB

README.md

File metadata and controls

213 lines (172 loc) · 10.7 KB

For the third time, the European Space Agency's Advanced Concepts Team (ACT) presents in cooperation with the The Genetic and Evolutionary Computation Conference (GECCO) the Space Optimization Competition (SpOC). Look forward to three challenging problems placed in an futuristic space mission scenario.

SpOC 2023

SpOC 3: Orbital Megastructures

The year is 2287 and you wander through the tunnels of Lunar City. Passing by you favorite kiosk, which is the only one on the Moon that still sells newspapers made out of real paper, an article in The Earth Observer catches your eyes. You grab a copy and a cup of 100% Quetelet-Crate-Bean coffee.

Winners of the Orbital Megastructure competition announced.

After decades of negotiation, two of the most ambitious projects that humanity has ever conceived are finally scheduled to commence next year. The Golomb Ruler Advanced Interferometric Lens (GRAIL) and the Orbital Assembly of Self-organising Interstellar Spacecraft (OASIS) projects have been selected among more than three dozen candidates competing for the immense amounts of funding. Despite the fierce competition, the unanimous opinion is that the sheer ambition and scope of cooperation on these projects will elevate humanity to a new level as a species.

Although the winning projects could not be more different, they are inexorably linked. The veteran GRAIL project (the oldest of all proposals, first conceived in the distant 21st century), overcame severe criticism and convinced the judges with solid arguments for its usefulness. The GRAIL project became relevant overnight with the discovery of Gaia Gemina - the likely habitable 'Earth twin' planet located squarely in the middle of the Goldilocks zone around an inconspicuous Sun-like star in the constellation of Sagittarius. The GRAIL team promises that the project will be completed within record time using new materials and orbital stabilisation mechanisms, enabling astronomers to examine Gaia Gemina and its cosmic neighbourhood in unprecedented detail just a few years from now.

In contrast, OASIS is a latecomer to the project race. It represents the hope of humanity to become a space-faring species by sending the first crew of people to Gaia Gemina. Initially viewed as a practically impossible task, the construction of the multi-generational interstellar ship is now scheduled to begin in orbit around Earth at the same time as the GRAIL project. The ship - which has yet to be given a name - must be large and sophisticated enough to sustain several generations of pioneers on their way to Gaia Gemina.

Of course, both GRAIL and OASIS would depend on the outcome of the Graph Reduction Algorithm for Planetary-scale Hyperoptimisation (GRAPH) meta-project. Originally selected as the tool of choice for solving challenges such as those posed by GRAIL and OASIS, the pressure for this approach to work has increased manyfold with the completion of the multi-year negotiation marathon for funding GRAIL and OASIS. As the winning projects are scheduled to commence next year, it remains to be seen whether GRAPH was indeed the right choice.

Whatever the outcome, rallying behind these projects - not at the local or national but at the planetary level - is no small feat, and the sense of accomplishment and anticipation is already palpable. Needless to say, the stakes are high for all projects to succeed in the mission to guarantee the safety of the first intrepid interstellar explorers!

Competition Structure

SpOC 3.0 contains three distinct problems centered around a futuristic space mission. Starting from 1 April 2024, End of Day, Anywhere on Earth (AoE) you have three months to tackle these challenges to secure a spot on the leaderboard, i.e., until 30 June 2024, End of Day, Anywhere on Earth (AoE).

Detailed technical descriptions for the three challenges to be solved will be made available on the Optimise platform from the same date.

Challenge 1: Torso Decompositions

Torso Decomposition

Before we can start designing astonishing megastructures in the GRAIL or OASIS project, we need to arrange the building blocks suitably. These blocks, called nodes or vertices, are bound to specific pairwise interactions. Some blocks clearly belong to the same part of the megastructure, while others share no interaction. For an effective and energy-efficient structure setup, it is critical to identify an ordering of the nodes that indicates parts of the structure that can be assembled concurrently, and that encodes a global blueprint for recursively constructing the megastructure. At the Graph Reduction Algorithm for Planetary-scale Hyper-optimisation (GRAPH) meta-project, we develop the toolkit for exactly this task -- and we hope that engaged scientists and engineers like you will help us to complete the project in time!

Challenge 2: Interferometric Mission

Interferometric Mission

Welcome to the GRAIl project, where our sights are set on one vital mission: high-resolution interferometric measurements. These measurements are crucial to map distortions in space-time in order to effectively navigate through warps and folds. Without their high precision, spacecraft risk being lost in the fabric of space-time, never to return. Scientists figured out that a constellation of satellites can achieve the needed precision by forming a large synthetic aperture. Unbeknownst to them, there is an intricate hurdle that they still need to overcome: how to arrange the satellites such that they maintain a useful configuration over several orbits? It turns out that the configuration of satellites which results in the highest resolution is also the one which best follows the mathematical structure of Golomb patterns. Synchronizing these orbits and coordinating observations requires the collective expertise of scientists, engineers, and innovators like you!

Challenge 3: Programmable Cubes

Programmable Cubes

Welcome to the OASIS project, one of humanity's flagship research projects for in-space assembly of megastructures! Even way back in the 21st century, scientists envisioned ships and habitats to self-assemble in space from small components that can be easily deployed from Earth. Now this dream is finally getting closer with the programmable cubes technology: small cubes of different designs capable of forming complex and fully functional structures when interconnected.

The missing building block for this technology is an efficient and scalable algorithm for assembling structures. To solve this, we need the help of every student, scientist and engineer out there interested in contributing to the OASIS project! To test your algorithms, we generated prototype scenarios where you have to assemble two historical spacecraft from the early days of human space exploration: The International Space Station (ISS) and the James Webb Space Telescope (JWST). For final evaluation, we also give you a single assembly scenario of the planned multi-generational interstellar ship Enterprise. If successful, your algorithm might give us the missing puzzle piece for constructing this massive ship in orbit and boldly go where no one has gone before!

Guidelines and Rules

The competition will be hosted on the Optimise platform developed by the Advanced Concepts Team. Participants will need to register online on the platform, and solution entries will need to be submitted via Optimise for validation. While SpOC is organized in cooperation with GECCO 2024, it is not required to attend GECCO 2024 in order to participate in SpOC.

  • Your objective is to propose and implement metaheuristic algorithms to solve the proposed optimisation challenges.
  • In order to validate your solutions, we will provide you with Python validation code for each of the three challenges. This code includes problem definitions in the Pygmo user-defined problem (UDP) format, examples of solutions, and visualisation tools.
  • You have until 30 June 2024 to submit your entries via the dedicated portal Optimise.
  • Please comply with our basic code of honour. The ACT reserves the right to exclude users from the competition if they abuse the evaluation system.

Scoring and Winner Selection

This year, SpOC implements a 2-out-of-3 scoring scheme that, simply put, means that for the overall competition only two challenges are taken into account. In detail, you will obtain a local score $s_i$ for each of the three challenges computed via the the rules below, then your global score is $g=\max(s_1+s_2, s_1+s_3, s_2+s_3)$. The global score defines your place on the overall SpOC leaderboard and the winners of SpOC 3.

Each challenge $i=1,2,3$ contains three problems (easy, medium, and hard). Every problem has its own leaderboard (visible on Optimise) that ranks participants according to the objective of the challenge. The top ten ranks on the leaderboard of the easy instance get $e_i=10,9,8,\dots,1$ points, similarly $m_i=\frac{4}{3}10,\frac{4}{3}9,\dots,\frac{4}{3}$ points for the top ten teams on the medium problem, and $h_i=(\frac{4}{3})^2 10,(\frac{4}{3})^2 9,\dots,(\frac{4}{3})^2$ points for the hard problem. The score $s_i$ for the challenge then simply is the sum $s_i=e_i+m_i+h_i$.

We wish all participants the best of luck and are excited to see what you accomplish!

Timeline

timeline

The Space Optimization Competition starts at the first of April, End of Day, Anywhere on Earth (AoE):

countdownmail.com

NOTE: The submission portal remains open after 30 June 2024, End of Day, Anywhere on Earth (AoE). Submissions received after that date will not be taken into consideration for the competition, but still appear on the leaderboard.

Contact

Our primary means of communication with competitors will be the Discussions feature on this repository. Please use it to ask any questions you may have about the challenges or to exchange information with us. We will do our best to respond to your questions in a timely manner.

If you encounter a bug in the code, please use this repository's Issues feature to report it.