29 Postdoctoral Fellowship at KTH Royal Institute of Technology, Stockholm, Sweden

A fully funded postdoctoral position is available in a new robotics research group at KTH that explores how to adapt recent advances in AI/ML to improve the capabilities of real (fielded) autonomous robots acting in real-world workplaces and public spaces (mapping, planning, interaction). Join us to develop the next generation of AI-enhanced autonomy stack for fielded robots in complex real-world environments. We have a Boston Dynamic’s Spot with the Spot Arm. All work is expected to be validated in real robot experiments.

The successful candidate will perform research in artificial intelligence and machine learning applied to music beyond 12-tone equal temperament. How can the combinatorial space of melodic and chordal possibilities be efficiently searched for a user? How can a physical device or controller be rendered to make accessible these possibilities? The candidate will contribute to the development of music pedagogy for AI moving beyond 12-tone equal temperament, as well as the organization of workshops, conferences, festivals, and concerts. The candidate will also contribute to the maintenance of the computational resources of the project.

The 5-year research project, Music at the Frontiers of Artificial Creativity and Criticism (MUSAiC, ERC-2019-COG No. 864189), analyzes, criticizes and seeks to fundamentally broaden the transformation of three interrelated music practices by artificial intelligence (AI): 1) listening, 2) composition and performance, and 3) analysis and criticism. This two year postdoctoral position focuses on the musicological study of music generated by or with AI. How appropriate are current musicological practices for the analysis of such work? What new methodologies are needed? What does it mean to define a ”canon” when it comes to music generated by machines? How can a ”canon” be constructed? What is the use of studying music that by and large only machines will ever listen to? Does the world need a new kind of musicology?

Your role as a researcher in visual optics will mainly be to work on existing research projects about the effect of optical errors on quality of vision in the central and peripheral visual field. This will include the maintenance and update of an adaptive optics visual simulator, programming and setting up psychophysical routines, recruiting subjects, performing measurements, analyzing the relationship between image quality and vision, as well as documenting and presenting the research results. The work will be collaborative and include supervision of students and co-supervision of PhD-students. There are also possibilities to initiate and lead new research projects within visual optics as well as to gain teaching experience both on undergraduate and graduate level. Available funding offers the possibility of extending the position beyond the first year after evaluation. 

The division of Biophysics is located at Science for Life Laboratory, where the team develops some of the world’s most used tools for modelling biological systems with particular applications to membrane proteins and techniques for drug design. We are looking for a researcher to work on simulations of ligand-gated ion channels based on new data från cryo-electron microscopy, in particular to understand how channels open and are allosterically modulated based on binding of lipids and other small molecules. The work will also involve collecting and interpreting new cryo-EM data.

The division of Biophysics is located at Science for Life Laboratory, where the team develops some of the world’s most used tools for modelling biological systems with particular applications to membrane proteins and techniques for drug design. We are looking for a researcher to work on programming, maintenance, steering and project leadership within the development of the software code GROMACS for molecular dynamics simulations. The duties involve development of methods to apply simulations, in particular to membrane proteins, programming in C++ and for accelerators, as well as taking responsibility for organizing meetings and leading work packages with EU projects. The work will include collaboration and conference travel, in particular in Europe.

The AICell Lab (https://aicell.io) in the department of Applied Physics at KTH and Science for Life Laboratory is a new research group funded by the Data-driven Life Science Fellows program. The group, led by Wei Ouyang, focuses on building AI systems for data-driven cell and molecular biology. To expand the team, we are now opening a position for a highly motivated researcher. The researcher will play a key role in our ongoing work on AI-powered computational microscopy and whole-cell modeling. This is a unique opportunity to contribute to cutting-edge research in the field of data-driven cell and molecular biology and be part of an enthusiastic and dynamic team.

We are looking for a postdoc to work on algorithm development for distributed decision-making, optimization and machine learning, to strengthen our research environment in machine learning, communications, privacy and secrecy. As a postdoc your primary goal is to progress toward becoming a fully independent researcher, both through own independent work and in collaboration with other researchers at the department. The position will be placed with the Division for Information Science and Engineering at the EECS School.

Deep networks have been repeatedly proven to benefit from large amount of labelled training data.  Collecting high-quality labelled data, at Google scale, is prohibitively laborious and expensive for most societal bodies and industrial sectors. In this project, the postdoc candidate will work on unifying methods for undersupervised learning which includes semi-supervised, weakly-supervised, and noisily-supervised learning techniques. The postdoc will be hired at Hossein Azizpour’s group at the division of robotics, perception and learning and the project is in collaboration with a twinning postdoc at Aalto university in the group led by Juho Kannala. The postdoc position is funded by WASP (wasp-sweden.org) and is anticipated to involve two concrete applications in medical image analysis and autonomous driving.

Graph generative models (GGM) are  gaining significant interest in different application domains. They are commonly used to model social networks, knowledge graphs, medical data and bio/chemical networks. Large pretrained models like LLMs dominate already the generation of text and inspire the training of equally large graph pretrained models for respective domains. Recently multi-modality in LLMs is becoming popular with impressive results in the area of image, video and sound generation. On the other hand GGMs can also benefit from such approaches in domains where graph generation is needed including bio/pharma/medical applications (i.e. medical graphs/protein generation), electrical or communications networks, social networks etc. Thus, the work area for the post-doctoral researcher will be to design and innovate in this area with contributions to graph pre-trained models and in multimodal graph encoders and graph decoders. The issues of graph pre-trained strategies will be part of the research. Different models of generators will be considered including auto-regressive, one shot (i.e. GANs,  GAEs) or diffusion based ones.

Hydro power plays a central role in the energy transition in Sweden towards becoming carbon-neutral by 2045. To achieve sustainable power generation, the safety of hydro power facilities is the prerequisite. In a changing climate, hydraulic structures are experiencing higher risks in combination with aging issues. Modelling of flow and structural interaction plays a significant role in the assurance of safe spillway operation. This project aims to, by means of CFD, laboratory experiments and machine learning technique, enhance our understanding of failure mechanisms of stilling basins and develop physically based models for stability analysis. It also seeks to disseminate the research findings for applications in upgrades and operations of hydraulic structures.

The advertised post-doctoral position is within an ambitious, European Research Council (ERC) Consolidator Project titled “ULTRA”, led by Prof. Dejan Kostic. In this 2 Million EUR project, we want to dramatically change the way Internet Services are constructed. Special emphasis will be paid to low latency, resilient frameworks for machine learning (including distributed and federated learning). Examples of other services to be built include advanced highly streamlined server software, Network Functions Virtualization (NFV) chains at the true speed of the underlying hardware, intra datacenter load balancing, agile and efficient geo-distributed data storage systems that meet their Service Level Objectives.

In recent years, nanocelluloses have emerged as very interesting building blocks in new biomaterials. Cellulose nanofibrils (CNF), with a cross section of 5×5 nm and a length over 1 µm, can be extracted from different types of plants, but our focus has been on extracting them from the Swedish forest raw material. The anisotropic fibrils have a very high longitudinal/transverse ratio (>200) and a very high E-modulus (>130 Gpa), which makes them highly interesting as reinforcement materials in composites, among other things. It has also been shown that the fibrils can be very interesting as functional chemicals in e.g. papermaking, especially if they are functionalized with new reactive groups or by combining them with other chemicals/nanoparticles using the well-known Layer-by-Layer (LbL) technology where alternating cationic and anionic polyelectrolytes/nanoparticles are combined to achieve process and/or product improvements.

The advertised post-doctoral position is within an ambitious Digital Futures ISPP Project titled “Smart Predictive Maintenance for Pharmaceutical Industry (SMART)”, which is a collaboration between KTH and AstraZeneca. In this project, you will focus on developing digital twins, using data-augmented physical models, and integrating these with optimization methods for operations analysis and predictive maintenance. The position is at the Department of Mathematics in the Division of Optimization and Systems Theory, and you will be supervised by Jan Kronqvist (KTH) and Jeremy Vachier (AZ). The other 2 postdocs will focus on closely related topics. In the SMART project, you will work on-site in the AstraZeneca manufacturing facilities in Södertälje ~50% of the time with a goal to start in October 2023. This will afford you the unique opportunity to gain firsthand experience in an industry setting while also contributing to research of others in your cohort. Ultimately this should culminate in multiple prospects for co-authorships and patent applications, while delivering a minimum viable product (MVP) that increases the overall equipment efficiency (OEE) of the lines. Please see openings for the other SMART-Project here and here

 

(18) Postdoctoral Fellowship Position

Postdoc Fellowship Position summary/title: Postdoktor inom digitalisering av fastighetsdata

Deadline : 2023-06-07

View details & Apply

 

We envision a mobile cyber-physical system where people carrying mobile sensors (e.g., smartphones, smartcards) generate large amounts of trajectory data that is used to sense and monitor human interactions with physical and social environments. We seek to develop a causal artificial intelligence (AI) methodology to analyze and model human mobility behavior dynamics (decision-making) using individual travel trajectory data, and importantly develop the causal diagrams of human mobility behavior under disturbances that could help design effective strategies for sustainable and resilient urban mobility systems. We are looking for a highly-qualified and motivated postdoc in getting hands-on research experience in the area of Transportation Behavior and Artificial Intelligence. The postdoc will join a research team at the Division of Transport Planning. The postdoc develops the causal AI methodology and mobility behavioral graph under interventions, works closely with the researchers specialising in mathematics, causal inference, machine learning, data analytics and transportation science.

The advertised post-doctoral position is within an ambitious Digital Futures ISPP Project titled “Smart Predictive Maintenance for Pharmaceutical Industry (SMART)”, led by KTH and AstraZeneca.  In this industrial post-doctoral project, you will develop, evaluate, and implement novel methods to gather manufacturing data in a production environment. You will use your expertise in machine-sensor technologies, IoT, and operation condition monitoring to carry out your research.  In the SMART project, you will work on-site in the AstraZeneca manufacturing facilities in Södertälje ~50% of the time with a goal to start in October 2023. This will afford you the unique opportunity to gain firsthand experience in an industry setting while also contributing to research of others in your cohort. Ultimately this should culminate in multiple prospects for co-authorships and patent applications, while delivering a minimum viable product (MVP) that increases the overall equipment efficiency (OEE) of the lines. Please see openings for the other SMART-project positions here and here

The advertised post-doctoral position is within an ambitious Digital Futures ISPP Project titled “Smart Predictive Maintenance for Pharmaceutical Industry (SMART)”, will be led by KTH and AstraZeneca. In this project, you will focus on human-centric technological interventions in order to inform the creation of sensor networks and the parametrization of machine learning plant maintenance predictive models. This project will be led by Associate Professor Mario Romero. You will lead the creation of peripheral visualization interfaces to enhance plant operators with the information generated by the sensor networks and predictive systems. In the SMART project, you will work on-site in the AstraZeneca manufacturing facilities in Södertälje ~50% of the time with a goal to start in October 2023. This will afford you the unique opportunity to gain firsthand experience in an industry setting while also contributing to research of others in your cohort. Ultimately this should culminate in multiple prospects for co-authorships and patent applications, while delivering a minimum viable product (MVP) that increases the overall equipment efficiency (OEE) of the lines. Please see openings for the other SMART-project positions here and here

You will contribute to the development of methods for protein design through close interaction with leading partners at KTH and Chalmers at the forefront of biocatalysis and synthetic biology. The research will then be presented at international conferences and in international peer-reviewed journals. The postdoctoral duties include:

Ion channels are essential for life across organisms and a sophisticated understanding of their function is imperative for basic, biomedical and applied research. The overarching goal of the proposed project is to develop ion channel activators and inhibitors. Previously unmet specificity needs will be achieved by targeting subtype- and conformational-specific features, thereby limiting side effects. The focus will be on Kv7 channels linked to a range of severe neurological disorders, including epilepsy, sleep disorders and autism. The postdoc will carry out molecular dynamics simulations of these channels bound to various modulators, and integrate the results within a collaborative project with four other principal investigators carrying out cryo-electron microscopy, electrophysiology, behavioral experiments and machine learning.

The commercial vehicle industry is now undergoing a transformation from being a supplier of trucks, buses and engines to a supplier of complete and sustainable transport solutions. In in order to decrease the emissions of carbon dioxide from vehicle transports it is necessary to increase the share of renewable fuels in diesel. An increase of these so-called drop-in fuels can however lead to solubility problems and the formation of soft particles. These soft particles can give deposits in the fuel injection equipment and cause drivability problems. To be able to study and understand the mechanism behind internal diesel injector deposits (IDID) is important the increase use of drop-in fuels. KTH together with Scania and Scania Cummins XPI is studying this phenomenon in an ongoing project.

The position is part of a collaborative project with the Division of Wood Science and Engineering at Luleå university of technology, whose overall goal is to develop theory and algorithms for image-guided optimisation of the sawline. Specific tasks is to develop and implement physics aware deep neural networks for tomographic 3D image reconstruction with partially unknown acquisition geometry. Prototypes will be implemented as components in ODL (http://github.com/odlgroup/odl) that offers a seamless coupling between frameworks for tomography (ASTRA) and deep learning (PyTorch).

Cellulose and hemicellulose (and in addition lignin) are simple with regard of their atomic composition yet display an incredible variation of properties and behavior. As material they often contain water and different aspects of their macroscopic and molecular behavior depend on the water content on a manner that is neither trivial nor understood in all detail. A postdoctoral position is now available for studying the fundamental role of water in cellulosic materials both in their native state (the cell wall) and as processed material incl composites. The main experimental method is going to be solid-state NMR spectroscopy supported by scattering experiments and molecular dynamics simulations in cooperation with other researchers and research groups within the Wallenberg Wood Science Center (wwsc.se).

We offer a postdoc position for two years. The postdoc candidate will contribute to the development and implementation of mathematical models for the simulation of the heat transfer process in shallow geothermal systems for heating and cooling of buildings. The research will include both physical modeling as well as data-driven modelling approaches based on machine learning. The objective is the development of predictive models suitable to inform the operations of a controller for geothermal applications. The work is part of an international project funded by GEOTHERMICA Era-Net, in cooperation with the network Joint Programming Platform Smart Energy Systems. The overall goal of this international project is the development and testing of an innovative controller device to improve the operation of borehole fields coupled with heat pumps, thereby reducing their energy consumption, costs, and CO2 emissions, and take advantage of the full potential of the technology.

We are looking for a PostDoc to apply machine learning methods (ML) to modeling of severe accident phenomena in nuclear power plants and risk analyses. Designs of currently operating plants, modern small modular reactors and liquid metal cooled reactors will be of interest. The ML tools will be developed and trained to predict evolution of the accident progression in time with high computational efficiency. The trained ML models will be used to solve such problems as (i) assessment of the risk of releases of radioactive products to environment, (ii) development of severe accident simulators and operator support systems to minimize the consequences of the accidents. The work will be a part of national and international collaborative projects. You will be a part of research team with several PhD students and senior researchers.

In this postdoctoral position, you will conduct research in the development of new methods for balancing unstable hydrofoil boats based on a new monofoil concept for hydrofoils developed in the research at KTH. The challenges lie in optimizing the use of available sensors and applying creative new solutions using e.g. MPC, LQR and AI for maximum robustness and performance in stabilization. The research is closely related to the development of control strategies for unstable aircraft and spans areas such as theoretical modeling, development of new algorithms, implementation in simulators as well as on own boats with full-scale testing at sea.