Thesis Worker - Towards Trustworthy Agentic Software Development for Automotive C++ Systems

30 hp - Towards Trustworthy Agentic Software Development for Automotive C++ Systems 

Introduction 

Thesis work is an excellent way to get closer to Traton and build relationships for the future. Many of today's employees began their Traton career with their degree project. 

Background 

Artificial intelligence is rapidly transforming software engineering. New agentic AI systems can independently interpret requirements, implement functionality, generate tests, review code, and deliver complete software changes. This creates an opportunity to fundamentally rethink how software is developed. 
 
For automotive software, however, the key challenge is not whether AI can generate code, but whether engineers can trust it. Safety-critical systems require evidence that software is robust, maintainable, secure, traceable, and compliant with established engineering standards. Future automotive software development will require organizations to justify trust in AI-assisted development processes while maintaining compliance with regulatory and engineering requirements. Today, there is no widely accepted approach for evaluating and communicating confidence in software produced through agentic development workflows. 
 
This thesis offers an opportunity to explore one of the key unanswered questions in industrial AI adoption: how can we build justified confidence in AI-generated software for automotive applications? The work sits at the intersection of software engineering, AI, quality assurance, safety, and systems thinking, with the potential to influence how future automotive software is developed. 

Objective 

Investigate which mechanisms, quality gates, assurance techniques, and evaluation approaches can contribute to building confidence in software developed through agentic AI workflows for automotive applications. The thesis will develop and evaluate a methodology, framework, or prototype that provides evidence for assessing the trustworthiness of AI-generated automotive software 

Job description 

The student will conduct a literature review, analyze factors influencing trust in AI-generated software, investigate suitable assurance mechanisms, quality gates, or confidence metrics, develop a prototype or proof-of-concept, and evaluate the approach using representative automotive software development tasks. 
 
The exact scope will be defined together with the supervisor and may focus on areas such as software quality, verification and validation, cybersecurity, functional safety, traceability, explainability, trust metrics, or assurance frameworks. Rather than focusing on code generation itself, the thesis aims to answer the broader question of how organizations can responsibly adopt agentic software development in safety- and quality-critical environments. 

Education/program/focus 

M.Sc. students in Computer Science, Software Engineering, Artificial Intelligence, Embedded Systems, Cybersecurity, Functional Safety, or related fields. 

Number of students: 1-2 

Start date for the thesis work: According to agreement 

Estimated time required: 20 weeks (30 hp) 

Contact persons and supervisors 

Alexander Axelsson 
Associate Senior Engineer, TRATON Group R&D 
Alexander.x.axelsson@scania.com 

Christofer Bitzekis 
Unit Manager 
christofer.bitzekis@scania.com 

Application:

Your application must include a CV, personal letter and transcript of grades. 

A background check might be conducted for this position. We are conducting interviews continuously and may close the recruitment earlier than the date specified.