Master's thesis: Agentic Digital Twins for Energy-Efficient Software-Defined Vehicles

Background

The transition towards Software-Defined Vehicles (SDVs) is transforming the automotive industry into a software-intensive, highly connected ecosystem. Recent advances in Artificial Intelligence (AI), digital twins, and data-driven engineering enable continuous learning from vehicle development processes and real-world fleet operations. By leveraging predictive analytics, adaptive learning, and AI-assisted engineering approaches, vehicle manufacturers can optimize vehicle’s energy consumption throughout the lifecycle of SDVs.

A promising concept in this context is agentic digital twins, where AI agents interact with digital representations of vehicles and their environments to analyze system behavior, propose optimizations, and support decision-making. Agentic digital twins may operate through external, internal, or distributed agency, offering different levels of autonomy and interaction between virtual models, software agents, physical vehicles, and human stakeholders. Understanding how these forms of agency can improve operational efficiency is a key research challenge for future SDVs.

Description

This thesis investigates how agentic digital twins can support the engineering and operation of SDVs. The overall goal is to develop and evaluate a collaborative framework that integrates multiple AI agents, digital twins, and human experts to identify, validate, and continuously generate potential strategies for optimizing vehicle energy efficiency.

The proposed framework will model both the problem domain (e.g., vehicle dynamics, battery systems, sensing and actuation, driving behaviors, and operating environments) and the solution domain, where agents autonomously explore potential improvements and assess their impact. Human engineers remain in the loop by reviewing, validating, and governing AI-generated recommendations, particularly for safety-critical, business-critical, and security-sensitive decisions.

The key use case focuses on energy-efficient SDVs, where agentic digital twins continuously analyze vehicle data and energy usage patterns to identify opportunities for improved battery utilization, charging strategies, and overall energy efficiency.

Key Responsibilities

The student will design and evaluate an agentic digital twin framework for SDVs, focusing on the following objectives:

• Model external, internal, and distributed forms of agency within agentic digital twins for SDVs.

• Design and evaluate a multi-agent system that analyzes vehicle operational data and software to identify optimization strategies for energy efficiency.

• Incorporate human-in-the-loop mechanisms for validation, governance, and decision-making support.

The expected outcome is a prototype agentic framework that demonstrates how multiple AI agents, digital twins, and human experts can collaboratively optimize SDV operation.

Qualifications

Candidates are expected to be enrolled in a master's programme in a field related to computer science and engineering at a Swedish university. Having already completed AI-related courses and/or gained work experience with AI/ML is an advantage. Knowledge of digital twins or system modeling is also an advantage.

Conditions

• Location: RISE, Gothenburg

• Some physical presence is expected.

• Applications are reviewed on a rolling basis; apply as soon as possible, but no later than October 31st, 2026.

• Starting date: January 2027.

• Credits: 30 ECTS

• Compensation: 10.000 SEK upon successful completion of the thesis.

Welcome with your application!

Contacts

Dr. Rodi Jolak +46 10 228 42 56

Dr. Efi Papatheocharous +46 10 228 43 36