About the project
Objective
A Lego-inspired design framework called SiLago (Silicon Lego) enables automation from the system level to ready-to-manufacture solutions for high-performance Edge AI applications. This framework bridges the gap between ease of use and performance by providing ASIC-comparable efficiency while achieving significantly improved energy efficiency—10X to 100X better than commercial off-the-shelf (COTS) solutions such as GPUs and FPGAs. The research project aims to enhance the SiLago framework to support comprehensive system-level implementation by addressing computation, storage, and interconnect requirements. These enhancements will enable SiLago to streamline the synthesis of complex applications, such as those required in industrial use cases. Finally, the improved system-level capabilities will be seamlessly integrated with the existing application-level synthesis flow, creating a unified, automated design process from applications to manufacturable silicon.
Background
The field of electronics and VLSI has driven transformative advancements in computing, enabling the development of increasingly powerful and efficient hardware systems. System architecture plays a crucial role in defining the structure and interaction of hardware components, ensuring efficient computation, storage, and communication. Despite these advancements, designing high-performance and energy-efficient hardware, such as ASICs, remains a complex, resource-intensive process requiring specialized expertise. The SiLago framework builds on these foundations, combining principles of VLSI, hardware modeling, system architecture, and design automation to provide a modular, automated solution for ASIC design.
About the Digital Futures Postdoc Fellow
Nooshin Nosrati completed her doctoral research in Digital Electronic Systems at the University of Tehran (UT). Her doctoral thesis was on hybrid reliability provisions in embedded systems with a focus on Computational Elements. Her research interests encompass hardware design and modeling, computer architectures, reliability and testability of embedded systems.
Main supervisor
Ahmed Hemani, Full Professor, Department of Electrical Engineering, KTH.
Co-supervisor
Artur Podobas, Associate Professor, Division of SCS, School of EECS, KTH.
About the project
Objective
To develop a mobile application that leverages advanced 3D pose estimation technology for tracking and analyzing a person’s movement outside traditional laboratory settings. This app aims to complement the clinician’s work by enabling remote monitoring and interaction between patients and healthcare providers, thus facilitating data-driven physiotherapy and rehabilitation sessions.
The benefits of such an approach include:
- Adaptive, personalized therapy to each patient’s progress and needs
- Progress tracking highlighting achievements and areas needing attention
- Assisting in correct execution by providing immediate feedback
Background
The need for innovative tools that offer quantitative insights into patients’ movements has become increasingly apparent, particularly for remote or home-based physiotherapy and rehabilitation. Traditional methods often rely on in-person assessments that may not fully capture the nuances of a patient’s progress or challenges.
An app that provides accurate, quantitative movement analysis can significantly enhance the clinician’s ability to tailor treatments, monitor progress remotely, and ensure patients perform exercises correctly while reducing the need for frequent in-person visits.
Status
The current app demo utilizes a single smartphone or tablet to capture a person’s 3D movement using the device’s depth camera capabilities. To ensure high accuracy and reliability, the app employs a machine learning model to refine and improve pose estimation based on data collected from a wide range of users. The current implementation primarily targets the lower extremities, focusing on walking. Efforts are underway to broaden the model’s scope to encompass a broader range of movements.
Crossdisciplinary collaboration
The project partners are Innovations Office Region Stockholm and Danderyd University Hospital.
Watch the recorded presentation at the Digitalize in Stockholm 2023 event:
About the project
Objective
The QB-ACT project aims to develop and evaluate an Internet-based psychological intervention integrating Question-Based Learning (QBL) and Acceptance and Commitment Therapy (ACT). It seeks to create an engaging, user-friendly platform for delivering therapy to improve mental health outcomes, reduce dropout rates, and enhance treatment adherence. By conducting a randomized controlled trial, the project will assess its effectiveness in addressing anxiety, depression, and overall well-being. Additionally, it aims to promote accessibility and scalability, providing an innovative alternative to traditional mental health care while supporting broader adoption through stakeholder dissemination and integration into healthcare systems.
Background
The QB-ACT project addresses critical challenges in mental health care: the growing demand for services and insufficient accessibility. Internet-based psychological interventions offer scalable solutions but often face issues like low engagement and high dropout rates. To overcome these barriers, the project integrates the evidence-based framework of Acceptance and Commitment Therapy (ACT) with Question-Based Learning (QBL), a proven methodology from e-learning. ACT fosters mindfulness and value-driven action to improve mental health, while QBL enhances engagement and learning through interactive, problem-solving techniques. By uniting these approaches, QB-ACT seeks to create a transformative digital platform, making therapy more accessible, personalized, and effective.
Crossdisciplinary collaboration
The proposed research project unites a multidisciplinary team with expertise spanning clinical psychology, Internet-based psychological treatments, large-scale e-learning technology, instructional design, online education, and the OLI Torus e-learning platform, ensuring a comprehensive and innovative approach to digital mental health interventions.
About the project
Objective
The project aims to develop a model-based decision-making tool for planning, designing and improving patient flows at the Karolinska University Hospital. The tool will include a prediction model, integrating clinical data, production data, and expert knowledge of the patient flows through the emergency department and the hospital. In addition, a network-based simulation model will be integrated to facilitate capacity and resource planning to improve patient care, decisions on handling and resource utilisation.
Background
Hospital emergency departments play a central role in healthcare systems and patient flow logistics throughout the system. Delivery of emergency care is resource and knowledge-intensive and requires close alignment with all other hospital functions and departments. Emergency healthcare is a complex system with high variability in patient characteristics, disease profile, processes and outcomes. This is a challenge for change work aimed at improving efficiency.
Little or no work has combined clinical and production data into process models to inform healthcare planning. Current models treat the emergency department as an independent leverage point without considering downstream bottlenecks in hospital wards. Overcoming these methodological limitations is important for informing hospital resource planning.
Crossdisciplinary collaboration
The project partners are Karolinska University Hospital, Region Stockholm, and KTH.
Watch the recorded presentation at the Digitalize in Stockholm 2023 event:
About the project
Objective
The EFFECT project aims to develop a digital twin of electrified construction site resources, processes, and their dependencies to evaluate the potential cost (efficiency) and benefit (emission reductions) of best-practice electrification of a construction site. Partners KTH, ABConnect, and Gordian, along with the City of Stockholm and 3rd parties PEAB and Northvolt, will use the digital twin to evaluate a construction site and, in a planned continuation project, optimize the steps of electrified construction operations using AI methods similar to those used in chess machines. The digital twin and its application will be developed and tested within the Kvarter Persika living lab, an urban renewal project of 1200 apartments in Södermalm, Stockholm.
Background
By 2050, the number of people living in cities will increase by 60% to 6.5 billion. City construction today is responsible for 23% of the global carbon emissions. Electrification is the de-facto technology for decarbonising our society, including city construction. However, due to the variability, non-linearity, and relatively long duration of new processes linked to electrification, we need more knowledge about the potential benefits and costs of electrified constructions and smart methods for planning and optimizing electrified construction operations.
Crossdisciplinary collaboration
In the EFFECT project, academia, research-based startups, the city and large industrial partners from the construction and energy industries will combine academic disciplines of control theory, simulation, optimization, AI/ML, and network communications to make future city construction more sustainable.
About the project
Objective
Our objective is to understand the digital innovation gap in the Swedish Water and Sewerage sector in order to increase the speed of digital transformation of the sector.
Specifically, our goals are to:
- determine and assess structural, institutional or capacity-related barriers at the sector level;
- identify enabling factors for adopting ICT and Digitalisation in municipal infrastructure asset management specifically applied to water and sewerage.
Background
Information-driven decision-making for asset management and maintenance of water infrastructure holds great potential for efficiency gains. Yet, the uptake of digital innovations appears to be slow compared to other sectors. As large-scale water infrastructures are exposed to transformation pressure from ageing assets, demography, societal digitalisation, security risks and resource scarcity, identifying the innovation barriers – but also enabling factors – will be crucial. Our findings will also be relevant for other infrastructure-oriented organisations yet to make the digital leap.
Our approach is interdisciplinary and based on the social sciences and humanities, with an orientation towards the field of science and technology studies (STS). Method-wise, we use a case study approach, focusing on contemporary and recent historical cases of innovation in ICT and digital technologies in the sector. With this approach, we can analyse long-term sector experience of innovation and change.
Photo credit: Horst Gutmann, https://creativecommons.org/licenses/by-sa/2.0/#
Crossdisciplinary collaboration
Our team of researchers covers engineering, industrial dynamics, STS and history (PI, Co-PI and a post-doc researcher). We collaborate with a range of industrial and societal actors in this project, notably with DHI Sverige AB, Kungsbacka kommun, and Kommunalförbundet Norrvatten.
About the project
Objective
To understand how generative AI tools can be used by staff and students in the context of higher education. The project addresses three areas:
- Evaluate how such tools can be used by students to improve their productivity and learning outcomes
- Characterise how the technologies can be used by academic staff to transform education and assessment practices
- Provide guidance to university leadership regarding the regulation of use of such tools as well as capacity building initiatives that should be taken
Background
The sophistication of the latest generation of AI tools far exceeds that of previous generations, and from an educational assessment perspective the output is both sophisticated, and hard to detect. The realistic nature of the output is a product of the complexity of the systems and the scope of the data upon which they have been trained. Like many tools before them generative AI will transform our approach to education.
Crossdisciplinary collaboration
The project combines Human Computing Interaction and Education research competence from KTH and SU to address societal and technological aspects of the integration of generative AI into educational practices. By taking a multi-disciplinary approach the team is able to explore in depth both technological and educational dimensions of the use of AI, helping to craft the educational experience of the future.