PhD Position in Computational Physics - ,

Department of Physics at Umeå University (www.physics.umu.se) conducts strong research in the areas of organic electronics, condensed matter physics, nanotechnology, photonics, theoretical and computational physics. 

The Department of Physics is looking for a PhD student in computational science with a focus on physics and machine learning. The position is for four years of full-time doctoral studies. Application deadline is March 4, 2022.  The position will open in 1 April 2022 (exact start date can be negotiated).

Project Description
The project explores different ways of representing physics with deep neural networks and applications of this in automation and simulation. Specifically studied is granular materials and multibody systems with non-smooth dynamics, which can represent robots and vehicles that move in unstructured rough terrain and manipulate their surroundings. With virtual environments, based on physics simulation and 3D computer graphics, it is easy to generate synthetic training data for large amounts of scenarios. With deep learning, models can be trained to perceive the nature of the environment, how it will respond to interaction, and learn strategies to control a robot or vehicle to perform various tasks efficiently and safe. The research questions concern how accurately the simulated physics needs to replicate the reality for trained models to be easily transferable (sim2real) and how virtual environments can be automatically generated based on real observation data in combination with a physics library (real2sim).

Another research direction involves the possibility of using deep neural networks to represent physics models, for example representing the dynamics of granular materials or articulated rigid body systems. Can this lead to simulation models that are far computationally efficient than the conventional equation-based models and discretization techniques? Will these models be easier to calibrate and integrate with other systems that also have neural network representation?

The doctoral student position is linked to the research group Digital Physics located at UMIT Research Lab, which brings together computationally oriented researchers in the areas of computer science, physics and mathematics. The project is linked to the research programs eSSENCE (the national program in e-science) and Mistra Digital Forest. This offers the opportunity to participate national postgraduate courses in computational science and to collaborate with research groups and companies with unique domain knowledge, datasets, and infrastructure for automation and simulation of robots and terrain vehicles.

The employment is expected to result in a doctoral degree and the main assignment for the doctoral student is thus to be active in research, study doctoral courses, and possibly participate in teaching of courses at undergraduate level. The financing time is limited to four years full time. Teaching and other departmental work (up to a maximum of 20%) can be included. The employment is limited to four years at full time or up to five years if teaching and other departmental work is performed. The salary is fixed according to the established salary level for doctoral students. 

Competence requirements
To be admitted for studies at third-cycle level the applicant is required to have completed a second-cycle level degree, or completed course requirements of at least 240 ECTS credits, of which at least 60 ECTS credits are at second-cycle level, or have an equivalent education from abroad or equivalent qualifications.

To fulfil the specific entry requirements to be admitted for studies at third-cycle level in computational science and engineering, the applicant is required to have completed at least 90 credits in computational science and engineering courses, of which at least 30 credits shall have been acquired at second-cycle level. Computational science and engineering courses refers to courses with major quantitative, statistical or computing science elements, such as courses in computing science, mathematics and mathematical statistics, and physics, for examplesuccess  a master degree in physics, engineering physics, or equivalent. Applicants should have knowledge of computational physics or computational modeling and programming skills in C/C++ or Python. Experience of high-performance computing, control systems, robotics, or machine learning will be seen as an advantage. The candidate must be highly motivated and have the ability to work independently as well as a part of the research group. The applicant is required to be fluent in both oral and written English.

Application
The application should include the following:
1. A personal letter with a brief description of qualifications, research interests, and why you are interested in the position.
2. Curriculum vitae.
3. Certified copies of relevant degree diploma(s).
4. A list of university courses with grades. Note that for international applicants the grading system should be explained in brief.
5. A copy of master thesis and publications (if any).
6. Contact information of three reference persons.

Information
For more information contact Martin Servin, email: martin.servin@umu.se, tel: +4690-786 6508
Applications must be submitted via e-recruitment system Varbi no later than 28 February 2022.

We look forward to receiving your application!