MSc Thesis: Adapting a Car-Trained Vision-Language-Action Model for Truck Driving in Simulation

Location: Göteborg, SE, 417 15

MSc Thesis: Adapting a Car-Trained Vision-Language-Action Model for Truck Driving in Simulation

Location

Göteborg, SE, 417 15

#job-location.job-location-inline {

display: inline;

}

Position Type

Student

Permanent/Temporary

Temporary

Transport is at the core of modern society. Imagine using your expertise to shape sustainable transport and infrastructure solutions for the future. If you seek to make a difference on a global scale, working with next-gen technologies and the sharpest collaborative teams, then we could be a perfect match.

Background

The development and validation of Advanced Driver Assistance Systems (ADAS) and autonomous driving functions require large amounts of representative traffic data covering diverse traffic situations, environmental conditions, and road users. Collecting such data in the real world is costly and time-consuming, and certain critical or rare situations can be difficult to capture.

Virtual traffic environments and simulation-based testing offer opportunities to complement real-world testing by enabling repeatable experiments, controlled variation of traffic scenarios, and the generation of additional data for system validation.

Recent advances in sensor fusion, scene reconstruction, computer vision, and generative methods create new possibilities for reconstructing realistic traffic environments and generating parameterized traffic scenarios from real-world data.

In this thesis, we aim to explore these possibilities through two complementary research directions: reconstructing virtual traffic environments from multi-sensor reference data and generating parameterized traffic scenarios for ADAS validation. The specific focus can be adapted to the students’ interests and technical backgrounds.

Description of the Thesis Work

The thesis will investigate how real-world traffic data and simulation technologies can be used to support the validation of ADAS and autonomous driving functions. Depending on the selected direction, the work may focus on one of the following areas:

1. Virtual Traffic Environment Reconstruction

This direction focuses on reconstructing virtual traffic environments using multi-sensor reference data collected from trucks equipped with a reference sensor setup. The available data may include LiDAR point clouds, radar measurements, camera images, and IMU measurements.

The objective is to investigate how accurately the surrounding traffic environment can be reconstructed and how the reconstructed environment can support simulation and replay-based validation, including the replay of recorded sensor signals with software components in the loop.

Potential challenges include sensor synchronization, point-cloud density, occlusions, sensor coverage, and reconstruction fidelity.

2. Parameterized Traffic Scenario Generation for ADAS Validation

This direction focuses on extracting and interpreting traffic scenarios from real-world data and translating them into a formal scenario description language suitable for simulation-based validation. The objective is to investigate how recorded traffic situations can be identified, represented, and converted into structured, simulation-ready scenarios.

Potential areas of investigation include scenario extraction from recorded sensor data, interpretation of traffic participants and their interactions, mapping observed behavior to scenario parameters, and representing the resulting scenarios in a standardized scenario description language.

The two directions share the goal of improving simulation-based validation through more representative and flexible traffic environments. Depending on the selected approach, the work may also explore how reconstructed environments and generated scenarios can support synthetic data augmentation and more comprehensive testing workflows.

Potential Areas of Work

Multi-sensor data processing and fusion

3D scene and environment reconstruction

Virtual traffic environments and sensor simulation

Parameterized traffic scenario generation

Evaluation of reconstruction quality and scenario realism

Synthetic data generation and augmentation

Simulation-based ADAS validation, including false-positive and false-negative analysis

Suitable Background for the Student

We are looking for master’s students with an interest in autonomous driving, intelligent vehicles, and simulation-based validation.

Relevant academic backgrounds include

Machine Learning and Deep Learning

Computer Vision

Robotics and Sensor Fusion

Signal Processing and Sensor Modelling

Experience with Python, 3D data processing, simulation environments, or autonomous driving systems is beneficial.

Practical Information

Thesis level: Master’s thesis

Number of students: Two

Location: Gothenburg, Sweden

Language: English

Starting date: January 19, 2027

Application deadline: November 30, 2026

Are you interested in contributing to the development of next-generation validation methods for ADAS and autonomous driving? We look forward to hearing from you. Apply today!

We value your data privacy and therefore do not accept applications via mail.

Who we are and what we believe in

We are committed to shaping the future landscape of efficient, safe, and sustainable transport solutions. Fulfilling our mission creates countless career opportunities for talents across the group’s leading brands and entities.

Applying to this job offers you the opportunity to join Volvo Group. Every day, you will be working with some of the sharpest and most creative brains in our field to be able to leave our society in better shape for the next generation. ​We are passionate about what we do, and we thrive on teamwork. ​We are almost 100,000 people united around the world by a culture of care, inclusiveness, and empowerment.

Trucks Technology & Industrial Division hire team players who are ready to create real customer impact. Our decentralized teams work close to our custome