Thesis Project: 30 hp - Path Planning Using Generative Models

Introduction

Thesis work is an excellent opportunity to get closer to Scania and develop meaningful relationships for future career prospects. Many of today’s employees at Scania began their professional journey through their degree projects. By joining our team for your master's thesis, you’ll have the chance to work on cutting-edge technology while contributing to autonomous vehicle research, specifically focusing on path planning for trucks using state-of-the-art machine learning techniques.

Background 
Autonomous vehicles (AVs), particularly trucks, are a crucial part of the future transportation system, promising to increase efficiency, safety, and reduce environmental impact. Path planning, the process by which a vehicle decides the most optimal route to reach a destination, is one of the most critical tasks in AVs. Traditional approaches to path planning rely heavily on model-based methods, but recent advancements in artificial intelligence (AI) and machine learning have opened new possibilities. In particular, generative models offer novel ways to explore safe and efficient paths by predicting various outcomes and uncertainties.
Generative models, initially developed for image and text generation, have shown great potential in a wide range of applications, including time series prediction, decision-making. By utilizing generative AI, we can develop systems capable of generating possible paths for autonomous trucks in dynamic environments. These models can learn powerful strategies to iteratively improve path quality by balancing safety, compliance with traffic regulations, comfort, and travel time.

Objective 
The objective of this thesis work is to explore and apply generative models such as diffusion models for path planning in autonomous trucks. The goal is to leverage the generative capabilities of these models to propose a range of feasible paths for trucks while taking into account real-time constraints such as road conditions, obstacles, and traffic regulations. The project aims to develop an AI-driven system that can generate diverse, safe, and optimal paths for trucks in various environments, from highways to urban areas. Moreover, the thesis will develop techniques based on optimization and online learning to refine the generated paths over time, ensuring that the truck's performance improves as it encounters different scenarios.

Job description 
During the thesis period, the student will:

1. Conduct a literature review on generative models, and their applications in autonomous vehicle path planning.
2. Explore existing path planning methods for autonomous vehicles and identify areas where generative models could enhance performance.
3. Implement a generative model to generate candidate paths for autonomous vehicles, taking into consideration dynamic environments, road constraints, and real-time data inputs.
4. Develop methods and techniques to refine the generated paths and optimize them for safety, compliance with traffic regulations, and overall performance.
5. Test and evaluate the performance of the developed models in simulation environments.
6. Document the findings and deliver a final report that includes recommendations for future research and development in this area.

The project will involve significant programming, data analysis, and model training, with a focus on applying machine learning algorithms to real-world path planning challenges. Collaboration with other teams at Scania may also be required for access to simulation tools, traffic data, and autonomous vehicle platforms.

Education/program/focus
Master's student (final year) in:

• Computer Science/Machine Learning/Artificial Intelligence/Data Science
• Robotics/Autonomous Systems
• Any other related fields

Number of students: 1 Student 

Start date for the thesis work: January 2025

Estimated time required: 20 weeks (full-time)

Contact persons and supervisors
Mahmoud Selim (mahmoud.selim@scania.com)
Oscar Palfelt (oscar.palfelt@scania.com)
Sriharsha Bhat (sriharsha.bhat@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.
 

Thesis worker