About the project
Objective
The main objective is to develop a Content-Based Image Retrieval (CBIR) system using a large database of longitudinal brain Magnetic Resonance Imaging (MRI) of patients with dementia. By using artificial intelligence, the system will detect patterns and similarities in the longitudinal images, empowering healthcare professionals to predict treatment outcomes and deliver personalized care. Eventually, this tool aims to simplify decision making, improve patient care and make healthcare more efficient and cost-effective.
Background
As the global population ages, dementia diseases are becoming increasingly prevalent, currently affecting approximately 47 million individuals and imposing an economic burden of around $2.8 trillion. Innovative computer-aided diagnosis techniques, particularly CBIR, have been transformed by enhancing the retrieval of relevant images for patients with or at risk of dementia. Scientific evidence suggests that spatio-temporal patterns from longitudinal recordings can significantly improve outcomes in cross-sectional studies. However, progress in this field has been hindered by limited access to longitudinal databases, small dataset sizes, and ineffective analytical methods.
About the Digital Futures Postdoc Fellow
Félix received his PhD in Electronic Engineering from the Universitat Politècnica de València. His research has centered on applying artificial intelligence and signal processing techniques to biomedical data analysis. His thesis focused on developing a state-of-the-art preterm labor prediction system using Electrohysterography (EHG), successfully addressing challenges such as low incidence rates and limited data availability.
His work spanned the entire research process, from clinical data acquisition and signal preprocessing to the design of an automated decision-making system. Building on this foundation, he has expanded his proficiency in medical imaging modalities, and gained experience with modern deep learning architectures.
Main supervisor
Rodrigo Moreno, Associate Professor, School of Engineering Sciences in Chemistry, Biotechnology and Health (CBH), KTH Royal Institute of Technology
Co-supervisor
Chunliang Wang, researcher with Docent title at School of Technology and Health (STH), KTH Royal Institute of Technology
About the project
Objective
My research plan focuses on the second-order, online, and robust algorithms for decentralized machine learning, aiming to propose efficient algorithms with convergence guarantees and study the applications. For example, distributed training with multi-core processors enables faster and more efficient training of machine learning models; wireless sensor networks serve for smart buildings and automatic driving in decentralized manners; hospitals cooperatively study disease prevention and treatment while protecting patients’ privacy.
Background
Big data over geographically distributed devices is the new oil of the digital future. However, we cannot mine it within data centres due to privacy preservation and communication efficiency issues. Instead, we resort to learning over networks. My research plans to develop second-order, online, and robust decentralized algorithms with convergence guarantees. This plan perfectly fits the themes of Digital Futures: trust, cooperation, and learning.
About the Digital Futures Postdoc Fellow
Jiaojiao Zhang received a B.E. degree in automation from the School of Automation, Harbin Engineering University, Harbin, China, in 2015 and a master’s degree in control theory and control engineering from the University of Science and Technology of China, Hefei, China, in 2018. She received her PhD in operations research from the Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong (CUHK), Hong Kong, in 2022. She received the Hong Kong PhD Fellowship Scheme (HKPFS) in August 2018. Her current research interests include distributed optimization and algorithm design.
Main supervisor
Mikael Johansson, Professor, Division of Decision and Control Systems, KTH.
Co-supervisor
Joakim Jaldén, Professor, Division of Information Science and Engineering, KTH.
Watch the recorded presentation at the Digitalize in Stockholm 2023 event.
About the project
Objective
This project aims to propose the generalized design method and computational architecture of more than 20 kinds of nonlinear functions. By this they can be used in more algorithm acceleration without doing a lot of repetitive development work.
Background
As the foundation of the future development of smart society, the digital chips play a key role. Behind these digital chips is the hardware acceleration of a large number of algorithms. The essence of the algorithm is mathematical operation, the most complex mathematical operation is nonlinear function calculation. So the proposed research question is how to improve the universality of nonlinear function VLSI design without affecting the performance and efficiency? It is a challenge.
About the Digital Futures Postdoc Fellow
Hui Chen received a PhD from Nanjing University (NJU) in 2022, China. His major is information and communication engineering. Hui’s research interests include arithmetic circuits, integrated circuits for elementary functions and reconfigurable computing.
Main supervisor
Zhonghai Lu, Professor at the Division of Electronics and Embedded Systems, Department of Electrical Engineering, EECS, KTH.
Co-supervisor
Masoumeh Ebrahimi, Associate Professor, Division of Electronics and Embedded Systems, Department of Electrical Engineering, EECS, KTH.
About the project
Objective
This research will apply data-driven design to Climate Action by utilizing data from design, manufacturing, and the entire product lifecycle to learn how early-stage decision-making maps to downstream carbon emissions for complex systems.
Background
Artificial Intelligence has transformed the fields of Computer Vision and Natural Language Processing. Data-driven methods also have the potential to be a valuable tool in the fight against climate change, but not before such data is ready for computation. If data from product design, manufacturing, consumption, and retirement could be quantitatively represented for computation, then we could learn how to produce and consume more sustainably.
About the Digital Futures Postdoc Fellow
Haluk Akay completed his doctoral research in mechanical engineering at MIT. His doctoral thesis developed methods to represent textual design data for computation by extracting structured “what-how” information and evaluating designed systems using AI-based language modelling and design principles.
Haluk’s research interests lie in using design and data-driven methods to address complex climate change and sustainability problems. He also has experience in microelectromechanical systems (MEMS) fabrication and product design.
Main supervisor
Francesco Fuso-Nerini, Associate Professor, ITH, KTH.
Co-supervisor
Iolanda Leite, Associate Professor, Department of Robotics, Perception and Learning, KTH.
Watch the recorded presentation at the Digitalize in Stockholm 2023 event.
About the project
Objective
This research aims to design a new frontier for wireless extended reality (XR), achieved by intelligent reflecting surface (IRS)-aided terahertz (THz) communications, for providing reliable, low-latency and energy-efficient wireless XR services. Advanced machine learning algorithms will be exploited to develop intelligent, green, wireless XR solutions. The results are expected to apply to the wireless XR research and provide good insights for other fundamental wireless communication research areas.
Background
Extended reality (XR) is changing people’s lifestyles through the interaction of physical and virtual spaces. Wireless XR enables users to move freely and have a better quality of experience. However, wireless XR faces critical challenges such as high data rates, low interaction latency and limited power budget. Terahertz (THz) communications have emerged as a promising solution towards wireless XR due to its abundant spectrum. Meanwhile, intelligent reflecting surfaces (IRSs) can be deployed to compensate for the severe signal attenuation at THz frequencies. Although THz communications and IRS have been widely exploited for wireless communications, how to employ them in wireless XR is still an open question.
About the Digital Futures Postdoc Fellow
Chen Chen is a Digital Futures Postdoctoral Researcher at KTH Royal Institute of Technology. He received the B.E. degree from the East China University of Science and Technology, China, in 2018 and the PhD from the University of Sheffield, UK, in 2022. He was a Marie Curie PhD Fellow. From 2022 to 2023, he was a Postdoctoral Research Associate at the University of Liverpool, UK. His research interests include massive MIMO, wireless intelligence, wireless security, mmWave/THz networks, signal processing, and machine learning.
Main supervisor
Carlo Fischione, Professor, Division of Network and Systems Engineering, KTH.
Co-supervisor
Emil Björnson, Professor, Division of Communication Systems, KTH.
About the project
Objective
The research goal of this proposal is to develop a novel hierarchical control framework that enables multi-vehicle systems (MVS) to distribute, manage, and execute complex tasks in possibly dynamic and unstructured environments with safety guarantees and computational efficiency. For specified low-level navigation tasks, a unified theory will be developed devoted to distributed estimation and formation tracking control problems endowed with reactive collision avoidance abilities based on onboard local sensors (e.g., cameras or range finders). Decision-making mechanisms will be integrated to coordinate high-level navigation tasks to provide MVS with the capability to cooperatively specify the overall task and competitively allocate sub-tasks for each vehicle while ensuring efficient time and energy consumption.
The proposed framework will also enable a set (possibly the whole group) of agents to instantaneously choose and execute between different tasks reacting to real-time environmental changes in a prescribed mission. The innovative framework and algorithms will be developed with rigorous mathematical analysis and realistic simulations, potentially including engineered implementations that address practical urban challenges like search and rescue and intelligent transportation using aerial and sidewalk robots.
Background
Recent decades have witnessed the rapid expansion of autonomous robotic vehicles, such as self-driving cars, drones, autonomous marine vehicles, etc. They are envisioned as essential tools to venture into unsafe areas, enhance human sensory and manipulation abilities, or explore unstructured and possibly dangerous environments. Since a single autonomous agent typically makes it impossible to perform tasks in vast areas or complicated tasks that have to be decomposed into sub-tasks performed by multiple agents, both industry and academia have shown a great interest in the scientific area of networked autonomous vehicle systems, which are systems that can interact and coordinate with each other.
Although academic work on controlled multi-vehicle systems (MVS) has become ever-expanding recently, a large gap exists between current capabilities and required ones in real-world scenarios. For MVS to perform a wide range of tasks, autonomous navigation and control is a fundamental ability under which each vehicle should interact safely with neighbour vehicles and the surrounding environment. However, most existing control algorithms rely on full-state measurements, limiting their applicability to specific applications with suitably equipped experimental areas.
GPS signal is typically unreliable for practical scenarios involving tasks in urban canyons or congested environments. Hence, robust and computationally efficient distributed controllers and estimation algorithms must be designed based on onboard exteroceptive sensors (such as laser range finders, vision, and acoustic sensors). However, the documented results for robotic vehicles using onboard sensors have been limited to ad hoc scenarios. Moreover, the current practice of MVS is often conducting simple missions with restricted autonomy based on offline and centralized supervision and planning, assuming the environment is static and known, such as lighting shows via drones and delivering in warehouses via mobile robots.
About the Digital Futures Postdoc Fellow
Zhiqi Tang is a Digital Futures Postdoctoral Fellow at the Division of Decision and Control Systems of KTH Royal Institute of Technology, Sweden. From 2021 to 2022, she was a postdoctoral researcher with the Institute for Systems and Robotics. She was an Invited Assistant Professor in the Department of Electrical and Computer Engineering at Instituto Superior Técnico (IST) in Portugal.
She earned a double PhD in Automatic Control and Robotics from IST, University of Lisbon, Portugal, and I3S-CNRS, Université Côte d’Azur, France, in 2021. She obtained her B.S. in Electrical and Computer Engineering from the University of Macau, Macau SAR, China, in 2015. Her research interests focus on the estimation, control, and decision-making in multi-agent systems, with applications in Robotics and Transportation Systems.
Main supervisor
Jonas Mårtensson, Associate Professor, Division of Decision and Control Systems at KTH.
Co-supervisor
Karl H. Johansson, Professor, Division of Decision and Control Systems at KTH.
Michele Simoni, Assistant Professor, Transport Systems Analysis at KTH.
Watch the recorded presentation at the Digitalize in Stockholm 2023 event.
About the project
Objective
This research (3Dfire) aims to enhance forest fire detection and characterization through advanced remote sensing and AI techniques, facilitating improved management and mitigation strategies within forest ecoregions. Our AI algorithms will streamline the processing of extensive freely available remote sensing data, like Sentinel-1 and -2, to drastically reduce the time and costs associated with fire monitoring and prediction. Consequently, it will provide a comprehensive understanding of fire dynamics within the region. Additionally, “3Dfire” will identify and analyze the risk factors and drivers of fires in the Miombo woodlands, including the impact of extreme climate events and anthropogenic factors on fire severity, frequency, and duration.
The project’s outcomes will offer valuable insights to researchers and policymakers responsible for forest fire prevention, minimizing societal and ecological impacts, and promoting sustainable forest management practices. Ultimately, “3DFire” is pivotal in establishing a comprehensive global-scale perspective on AI for forest applications and transitioning towards climate-smart forest management.
Background
Miombo woodlands cover 270 million ha across southern Africa and are increasingly threatened by natural and anthropogenic forces. Despite their importance for biodiversity and 100 million forest-dependent people, this ecoregion receives little attention from the scientific community. The annual forest loss in just a part of the Miombo ecoregion was recently estimated to be more than half a million hectares. Fires are a primary cause of vegetation loss in Miombo, occurring approximately 50% more frequently than in other global ecoregions. They also trigger massive CO2 emissions in sub-Saharan Africa and threaten many ongoing sustainable forest management projects, such as Reducing Emissions from Deforestation and Degradation (REDD). The lack of accurate fire records challenges the estimation of CO2 emissions and evaluation of fire management activities in forests.
Although some dimensions of Miombo’s wildfires, such as frequency and severity, have been explored through coarse- and medium-resolution satellite time-series data from MODIS and Landsat, knowledge of the exact spatial extent, duration, and timing of the fires and their relations with effective driving forces is still lacking. However, fire regimes in miombo are dominated by small burn patches, particularly in the drier regions. Their limited spatial resolution makes it difficult to detect and characterize these small fires using the aforementioned satellites. Extreme climate events significantly elevate the probability of forest fires, while anthropogenic drivers affect all dimensions of fires. Meanwhile, extreme climate events and human pressures on forests are expected to increase as global temperatures rise and the population of southern Africa doubles by 2050. To address these limitations, this research will focus on characterizing and mapping various dimensions of fires, including the severity, frequency, timing, and duration of forest fires in miombo. We will leverage the time series data of fires derived from the Sentinel data and employ deep learning-based algorithms, particularly our developed residual attention UNet5 (RAUNet5).
About the Digital Futures Postdoc Fellow
Zeinab Shirvani is a postdoctoral researcher affiliated with the Division of Geoinformatics at KTH. She obtained a PhD in Remote Sensing/Cartography focusing on forest disturbances from TUD, Germany, in 2020. Before her current position, Zeinab was a postdoctoral researcher at the Swedish University of Agricultural Sciences (SLU) for two years. Zeinab has made significant contributions to remote sensing and machine learning throughout her research career, particularly in mapping woodland fires in tropical dry forests using RAUNet. Her expertise lies in applying geospatial artificial intelligence and remote sensing techniques to study forest disturbances.
Main supervisor
Yifang Ban, Professor and Director of the Division of Geoinformatics at the Department of Urban Planning and Environment at KTH.
Co-supervisor
Ulla Mörtberg, Professor and associate professor (docent), Department of Sustainable Development, Environmental Science and Engineering (SEED) at KTH.
Watch the recorded presentation at the Digitalize in Stockholm 2023 event.