Industrial PhD - foundation models & sensor data

Tackling real-world challenges in autonomous systems

We hope you are a talent who has recently finished your master's program or is about to graduate in the next few months! Or maybe you have one or two years of work experience. Perhaps you are looking for a PhD position at a top university while remaining close to the industry to understand real-world problems. Please continue reading if this is you.

The role

Scania is developing autonomous transport solutions and has recently announced autonomous trucks for mining applications. Widening the operational design domain (ODD) from confined areas to, e.g., transportation of goods on public roads is a significant undertaking. There will always be so-called long-tail problems, scenarios that are hard, or even impossible, to foresee and test the system against. The autonomous system must therefore be so generic that it can handle such unseen scenarios in the ODD. In recent years, foundation models have shown promise in enabling generalization and reasoning and a thriving research field is investigating different applications of these models, including autonomous driving.

One key aspect of autonomous systems is their sensing of the external environment. For autonomous vehicles, a sensor setup of, e.g., cameras, lidars, and radars is used. The sensor data must be processed in such a way that no important information is lost but it is still compressed to a manageable amount for the downstream tasks to be performed in a timely manner.

Scania and KTH are setting up a joint project, Multi-modal Open-set Perception for Safer Autonomous Trucks (MOPS), partly funded by Vinnova, where sensor data processing for and by foundation models will be studied in the context of autonomous driving. In this project, Scania will hire one industrial PhD student who will conduct research in this area supervised by KTH and Scania. 

The PhD student will be employed by Scania CV AB, and affiliated to the Division of Robotics, Perception and Learning, KTH Royal Institute of Technology. A PhD student position constitutes normally 80% academic work and includes coursework and research. The academic supervision is done by Olov Andersson, more information about his research and research group can be found here https://www.kth.se/profile/olovand

We are dedicated to an inclusive and diverse workplace, and you will be working alongside a diverse and talented group of developers, designers, and engineers who are committed to devising innovative solutions that redefine the way we deliver and learn.

Job responsibilities

• Research, write scientific papers, write PhD thesis.
• Take academic courses required by the PhD education.
• Work with applications of foundation models and other deep learning methods.
• Support and coach product development projects in proof-of-concepts validating approaches resulting from the research.
• Tech transfer of the research results to ensure long-term benefits at Scania, taking non-technical feasibility into account.

Your profile

• You hold, or will soon obtain a Master of Engineering degree (or equivalent) in applied mathematics, computer science, machine learning, robotics, or a related technical field. 
• Knowledge in deep learning.
• Preferably you have written a Master´s thesis on deep learning, and have solid theoretical knowledge of foundation models, sensor data, and vehicle autonomous systems.
• Strong coding skills and the ability to demonstrate results through running software on robotic platforms.
• Fluency in English, both in writing and speech.

If you consider this to be your dream job and feel that you have a fair share of what is mentioned above, please apply!

Our offer

We offer a healthy and inspiring workplace with respect for the individual, where you have great opportunities to shape your work and your future. Your personal and professional development is important, and we encourage and support it. We offer permanent relocation support for foreign applicants. 
 

For additional information

Thomas Gustafsson, Scania developer and project leader of this project, E-mail: thomas.gustafsson@scania.com
Mohammad Nazari, Scania developer and product owner, E-mail: mohammad.nazari@scania.com
Maria Linnarsson, Scania group manager, E-mail: maria.linnarsson@scania.com
Olov Andersson, academic supervisor at KTH, E-mail: olovand@kth.se

Application

Selections and interviews will take place throughout the application period. Please apply no later than 2025-01-27. 

Your application should include a CV, and degree certificate or marksheets. Instead of a cover letter, please answer the questions stated when you submit your application. A background check might be conducted for this position.

#LI-Hybrid