CS BSc Thesis Presentation 22 September 2026

Tuesday, 22 September there will be a bachelor thesis presentation in Computer Science at Lund University, Faculty of Engineering.

The presentation will take place in E:2405 (Glasburen).

Note to potential opponents: Register as an opponent to the presentation of your choice by sending an email to the examiner for that presentation (firstname [dot] lastname [at] cs [dot] lth [dot] se). Do not forget to specify the presentation you register for! Note that the number of opponents may be limited (often to two), so you might be forced to choose another presentation if you register too late. Registrations are individual, just as the oppositions are! More instructions for opponents are found here on the LTH thesis project page.

11:15-12:00 in E:2405 (Glasburen)

• Presenter: Sofia Lundgren
• Title: AI-driven automation and digitalization of construction control plans
• Examiner: Emelie Engström
• Supervisor: Simon Kristoffersson Lind (LTH)

A control plan must be created for most construction, demolition, and land works that require a building permit or notice. A control plan is a document stating which controls need to be completed during the project to comply with the Swedish Plan and Building Act. Manual creation and management of plans are time-consuming and tedious to maintain. This thesis implements a prototype web application to ease management and improve the integrity and traceability of control documents. Additionally, a RAG-based LLM integration enables automated control plan generation. The prototype was evaluated through comparative analysis against a professional reference plan and a general-purpose LLM (Gemini). Furthermore, feedback was collected from building permit officers and construction workers. Results show that RAG improves technical detail and terminology over generic models. Building permit officers stressed that the majority of plans must stem from a project-specific risk assessment. Construction workers also emphasized the need for a more flexible workflow that could support digital signatures and deviation logging. Ultimately, the prototype is most valuable if it were to be integrated into a larger, digital workflow rather than operating as a standalone generator.