Ingénieur Logiciel IA

Blagnac  - Stage (6 Mois)

À propos de Samares-Engineering

Samares-Engineering is a recognized expert in the field of systems engineering and more specifically in Model-Based Systems Engineering (MBSE). Samares-Engineering provides consultancy for design offices of large companies in different domains: Airbus, Zodiac Aerospace, Schneider Electric, Continental Automotive, Rockwell Collins... Samares Engineering also does research and provides training courses. It has strong relationships with high schools and research institutes in Toulouse: ISAE Supaero, Enseeiht, CNAM, IRT St-Exupery.

Description du poste

Amongst systems engineering activities, capture of stakeholder requirements and classification of those requirements into categories (functional, safety, installation…) are time-consuming activities. Requirements come into documents of various formats (excel, PDF, Word, image…) and it is necessary to parse those documents to extract the requirements and register them in a requirements database for future analysis. According to the requirement statement, systems engineers must choose the category that fits the best to classify the requirement. This classification is useful to prepare collaborative work on requirements: classification is used to distribute work (requirement validation and requirement analysis) to the right specialists (design, safety, certification, maintenance…). Samares-Engineering and Safran Seats are currently investigating an approach relying on Artificial Intelligence to simply and accelerate those two activities (capture and classification). They have defined an internship hosted by Samares Engineering to investigate this approach.


Goals:

  1. Validate the concept of pre-processing input data (textual) in order to automatically: Convert input media, identify requirements, classify Requirements, rate requirements narrative quality.
  2. Identify technologies to mobilize: Machine Learning, Natural Language analysis… , identify, compare, rate … identified technologies (preferably open-source)
  3. Develop a prototype of the pre-processing chain: Using the most promising technologies identified, producing data in a neutral format that can be injected in requirement management tools.
  4. Issue recommendations and a general design for “industrialized” development.
  5. Setup of demonstration and video: from previous steps you will build a demonstration and video showing how it works. You will be supported by whole team of Samares. Work will also be presented to Safran. If promising, results will be published to a journal or a conference and you will present it jointly with R&D engineer from Samares.
  6. Final report: synthesis, recommendations, and suggestions of new features to be implemented.

Description du profil

We are looking for interns with first engineering background and especially in Software Engineering to be able to develop a prototype, and curious about learning new methods and tools.


  • JAVA
  • Natural Language Processing
  • Python
  • Systems Engineering knowledge is a plus

Propulsé par