Master Thesis Work - Local Virtual CMDB for e-SBOM and Release Traceability

At ABB, we help industries outrun - leaner and cleaner. Here, progress is an expectation - for you, your team, and the world. As a global market leader, we’ll give you what you need to make it happen. It won’t always be easy, growing takes grit. But at ABB, you’ll never run alone. Run what runs the world. This Position reports to: R&D Global Unit Lead Your Role And Responsibilities Execution of this Thesis work will be done within ABB Automation PCP, a unit within ABB’s business area Process Automation. Automation PCP develops automation products - hardware and software - for monitoring, process control and optimization. We have automation products for all kind of industries. Within Automation PCP, we would like to continue the previous master's thesis on extended Software Bills of Materials (e-SBOMs). The next step is to investigate whether a locally hosted, retrieval-augmented large language model can provide an evidence-grounded 'virtual CMDB' for Configuration Status Accounting across distributed e-SBOM, development, release, and product-lifecycle information. Details:

• Period: TBD
• Number of credits: 30 ECTS
• Number of students for this thesis work: 2
• Location: Mostly on site - Malmo, Sweden

The thesis should use ABB System 800xA as the main industrial case and focus on a limited set of high-value questions spanning selected sources such as Azure DevOps, e-SBOM records, Dependency-Track, Windchill, source or artifact repositories, tests, and approved release documentation. Examples include identifying which artifacts and SBOMs belong to a release, tracing a component to builds and verification evidence, and detecting missing, stale, or conflicting information between systems. The virtual CMDB must be treated as a query and interpretation layer, not as the authoritative configuration record. Source systems, approved baselines, explicit links, stable identifiers, versions, lifecycle states, and ownership rules remain authoritative. Answers produced by the prototype should identify their sources, distinguish approved from non-approved information, expose conflicts or insufficient evidence, and avoid unsupported conclusions. A working proof of concept is expected. It could use e.g. scripts and REST APIs to ingest and normalize a manageable subset of configuration information, combine exact-identifier, keyword, and semantic retrieval where appropriate, and use a locally hosted LLM to answer prioritized questions with references to source evidence. New or changed configuration items should be incorporated through incremental ingestion and indexing rather than retraining the model for every update. Research questions to consider as input:

• Which Configuration Status Accounting questions provide the greatest value for managing e-SBOMs, software releases, and product traceability in the ABB case environment?
• To what extent can a locally hosted retrieval-augmented virtual CMDB answer these questions correctly and usefully across selected distributed sources?
• How do structured relationships, exact search, semantic search and hybrid retrieval affect retrieval quality, answer correctness and evidence groundedness?
• How effectively can the solution detect missing, stale, or conflicting configuration information and state when available evidence is insufficient?
• How can continuously added or changed configuration items be incorporated while keeping answers current without retraining the underlying model?
• Which governance and technical controls are needed to prevent generated answers from being mistaken for approved configuration or release evidence?

Expected result: The expected result is a master's-level thesis combining an industrial case study, design science, and experimental evaluation. Deliverables should include a prioritized taxonomy of Configuration Status Accounting questions, a documented architecture and information model, a working local prototype, and a controlled evaluation using verified reference answers. The evaluation should distinguish retrieval quality from final-answer quality and address correctness, source-groundedness, unsupported claims, conflict detection, update freshness, response time, manual investigation effort, practitioner usefulness, security, maintainability, limitations, and a roadmap toward industrial use. Qualifications For The Role

• ABB are looking for a pair of students for a thesis work at master’s degree level in Software Engineering, Computer Science, Machine Learning, Artificial Intelligence, Information Systems, or a related field.
• ABB are looking for students in the education phase where a thesis project is needed to finalize the education.
• It is preferable that at least one student has completed a course in software configuration management and understands configuration items, baselines, versioning, change control, configuration status accounting, and traceability.
• It is preferable that at least one student has completed a course in machine learning or artificial intelligence and understands model evaluation, data pipelines, retrieval methods, and responsible handling of model limitations.
• You understand DevOps and CI/CD concepts and have practical experience with scripting and REST APIs, for example using Python or PowerShell.
• You have general knowledge of Software Bills of Materials, software dependencies, lifecycle traceability, and the need for reproducible and auditable engineering evidence.
• English is ABB's defined corporate language. Understanding Swedish will simplify collaboration but is not mandatory.

More About Us We value people from different backgrounds. Apply today for your next career step within ABB and visit www.abb.com to learn about the impact of our solutions across the globe. #MyABBStory Recruiting Manager Ali Houmani, +46 72 461 43 24. For technical thesis related questions please contact Christer Wibom, +46 761369496. Positions are filled continuously. Please apply with your CV, academic transcripts, and a cover letter in English. We look forward to receiving your application! A future opportunity This role is part of our talent pipeline, which means it’s not currently open, but we’re always looking for curious minds, bold thinkers, and people who want to make an impact. Click Apply to express your interest and be considered for future opportunities that match your experience and aspirations. At ABB, we welcome people from all backgrounds and believe that diverse perspectives help us build a cleaner, smarter future. Apply today or visit https://www.abb.com to learn more about how we help run what runs the world.