Intern in the Mission Analysis Section, GPU Based Algorithms

Our team and mission

This position is based at the European Space Operations Centre (ESOC) - Darmstadt, Germany

Under the direct authority of the Director of Operations, the Ground Systems Engineering and Innovation Department is responsible for all the ground systems engineering and support activities required to support mission operations and implement innovation.
The Head of the Ground Systems Engineering and Innovation Department maintains and manages a full level of competences, technologies and services in all engineering disciplines related to ground systems, infrastructure, technology evolution and Engineering R&T and innovation. The department provides expert support in these areas to the Directorate of Operations for all current and potential missions and leads, in the role of architect, ground systems evolution and innovation implementation in the infrastructure. The Department is furthermore responsible for the overall technical standardisation coordination for the ground segment.

Candidates interested are encouraged to visit the ESA website: http://www.esa.int

Field(s) of activity for the internship

Topic of the internship: GPU Based Algorithms for Mission Analysis Applications

Graphic Processing Units (GPUs) are a computing technology that has recently gained prominence in the scientific community. While initially developed for graphical applications, its parallelization possibilities find several other uses, such as performing repeated operations on large data sets, running Monte Carlo simulations, and training Artificial Intelligence (AI) models. 

Despite significant progress has been made in the past few years, the use of the GPU technology for complex scientific computations is still limited to rather simple tasks, mainly due to the difficult code portability between CPU and GPU: the difference in the hardware itself requires to completely re-write existing software with a dedicated programming paradigm, which most scientists are not familiar with, preventing time and resource investments from communities well-acquainted with traditional CPU libraries. 

Nonetheless, a few application show that extending the GPU use to other scientific computing tasks has an enormous potential, and can allow scientific users to turn too-expensive analyses (for today's software) in almost real-time tasks. An example of such software is CUDAjectory, developed in the Mission Analysis Section at ESA, that performs orbital simulations using GPUs: while days (if not weeks) are required to run a few thousand-sized Monte Carlo orbital simulation using state-of-the-art combined Python/C++ code, the CUDA based trajectory propagation tool CUDAjectory can simulate millions of samples in just a couple of hours. Today, CUDAjectory is mostly used for the assessment of the compliance of mission trajectories with Planetary Protection and Space Debris Mitigation requirements: low impact probabilities with relevant celestial bodies need to be estimated with strict confidence intervals, often leading to hundreds of thousands to millions-sized Monte Carlo runs.

This internship aims to develop GPU algorithms to find orbits connecting two points in space-time (Lambert Problem) over a large and dense parameter domain, with the goal of generating parameter maps that aid Mission Analysts in the design and selection of spacecraft trajectories. While in the simple two-body format the Lambert Problem has established CPU codes and solutions, extending it to account for orbital perturbations make it a compute intensive task that would benefit from the adoption of the GPU technology. The successful implementation of a GPU Lambert Solver will be a significant technology demonstrator, proving the suitability of GPUs to solve a new class of problems. The concept resulting from this research can then be extended to solve some of today's most compute-intensive Mission Analysis problems, such as generating and optimizing Libration Point and low-thrust trajectories.

Behavioural competencies

Result Orientation
Operational Efficiency
Fostering Cooperation
Relationship Management
Continuous Improvement
Forward Thinking

For more information, please refer to ESA Core Behavioural Competencies guidebook

Education

You must be a university student, preferably in your final or second-to-last year of a university course at Master’s level and you need to remain enrolled at your University for the entire duration of the internship.

Additional requirements

The working languages of the Agency are English and French. A good knowledge of one of these is required. Knowledge of another Member State language would be an asset.