Requirements management is a critical and time-consuming phase in the software development lifecycle, especially in highly regulated sectors such as automotive, where compliance with strict standards (e.g., ISO 26262, ASPICE) is essential. The complexity of modern systems leads to an exponential growth of requirements, making it increasingly difficult to track them, verify consistency, and maintain traceability between different specification levels. Automating these activities can reduce the risk of errors, improve efficiency, and ensure higher software quality. In this context, the use of Machine Learning, Large Language Models (LLM), and black-box approaches offers new opportunities to optimize the requirements management process, minimizing manual effort and supporting complex tasks such as automatically linking system and software requirements or segmenting stakeholder requirements in management tools like Jama and IBM DOORS.
This thesis aims to explore and develop an innovative Python framework to address these challenges, contributing to the evolution of requirements management methodologies in the automotive sector.
Key Project Steps:
* Problem, methodologies and toolchain study
* Literature and next best alternatives analysis
* Implementing an automatic linking system between system and software requirements
* Developing an algorithm for the automatic itemization of stakeholder requirements
* Integrating the framework with requirement management tools such as Jama and IBM DOORS
* Evaluating system performance and comparing it with traditional methods
Main topic:
Development in Python – artificial intelligence algorithms.
Course of study and candidate requirements:
* Proficiency in Python and Machine Learning/NLP libraries (e.g., TensorFlow, PyTorch, scikit-learn)
* (Alternatively) MATLAB knowledge and optimization background
* Interest in software requirements management and automotive development methodologies
* Ability to analyze data and design AI-based systems
Location:
The internship can be carried out from one of the Kineton offices (Torino, Napoli or Reggio Emilia).
#J-18808-Ljbffr