Young Graduate Trainee for Machine Learning for Software Product Assurance

The mission of the Quality, Dependability & Product Assurance Support Division is to contribute to the success of ESA projects and activities by providing expert support and disseminating knowledge in the areas of responsibility and establishing quality management systems within ESA. The disciplines covered by it are system safety, dependability (reliability, availability, maintainability), quality assurance, quality management and software quality.

Quality assurance, safety and dependability engineering are primarily concerned with the development and implementation of methods, techniques and processes to achieve confidence for safe and reliable system design, manufacturing, operation and disposal.

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

The objective of this activity is to learn about a new technology, Machine Learning and its impact on software product assurance activities and related standards ECSS-Q-ST-80C and ECSS-E-ST-40C. This is not a theoretical exercise and includes hands-on to create a demonstrator for an engineering-use case or a development-phase case.

For this activity you will need to research the standards and take in an upcoming handbook on “Agile” and other soft skills such as Time Management following a “Personal Kanban” process. You will then follow online courses on the target topic of Machine Learning which will include the use of an ML computer Nvidia Jetson Nano using at least a couple of ML frameworks such as Tensorflow, Keras or PyTorch.

After these courses and after performing some additional research, you will identify the typical development lifecycles applied and the individual processes and activities along with their inputs and outputs. All these will be put into the context of our standards, matching them where possible and identifying those (if any) that do not have clear matches. You will also survey general knowledge (biased vs unbiassed data), best practices (known models that work in some cases) and accepted metrics (size of training data sets, etc.) that can be applied to ML development. All the abovementioned findings and knowledge will be included in a Technical Note.

Lastly, there will be a practical exercise where an engineering-use or a development-use case will be identified and implemented, following the Technical Note produced as a first output of this YGT opportunity. Depending on the use case selected, it will most likely be necessary to liaise with people in other ESA Departments/Directorates and possibly with industry or academic institutions. This activity may involve development of synthetic data sets for the training of the model, as gathering the necessary real inputs may be too difficult or even impossible.

Possible examples of potential applications:

• implementation of a Star Tracker with computer vision.
• predictors of operational incidents linked to the software quality metrics.
• code analysis with Machine Learning to identify hotspots for software problems.
• identifying problems at requirements level with the help of Machine Learning applications.

You should have just completed, or be in the final year of a university course at Master's level (or equivalent) in a technical or scientific discipline.

Basic knowledge or training on artificial intelligence or machine learning will be a definitive asset.

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.　

You should demonstrate good interpersonal skills and the capacity to work both independently and as part of a team. 　

During the interview your motivation and overall professional perspective/career goals will also be explored.