WASP Research
Artificial intelligence, autonomous systems and software are central to technologies that shape society, industry and research. WASP supports strategically motivated basic research within these three areas, strengthening Sweden’s long-term competence and the development of secure, reliable and useful intelligent systems.
The research in WASP is organized into four closely connected areas: AI/MLX, AI/Math, autonomous systems, and software. Most of the research is conducted through PhD student projects. There are also several initiatives involving large projects that span multiple research groups and bring together experts across disciplines.
AI/MLX
AI/MLX focuses on machine learning, deep learning and next-generation AI systems. The research develops methods that allow systems to learn from data, adapt to new situations and support decisions in complex environments.
Key topics include representation learning, multimodal data, reinforcement learning, sequential decision-making, transfer learning and learning from limited or continuously changing data.
Machine learning can support scientific discoveries, improve industrial processes, contribute to more efficient use of resources and create better decision support. For these technologies to be useful in practice, they must also become more reliable, explainable and adaptable.
AI/Math
AI/Math focuses on the mathematical foundations of artificial intelligence. The aim is to understand why AI systems work, when they fail and how they can be made more reliable.
The research includes verifiable methods, robustness, reproducibility, optimization and computational complexity, and connects AI with fields such as statistics, numerical analysis, topology, optimization and theoretical computer science.
Mathematical understanding is essential for developing AI that is more predictable, transparent and safe to use, especially in areas where errors can have significant consequences, such as healthcare, transport, engineering, science and digital infrastructure.
Autonomous systems
Autonomous systems research focuses on technologies that can perceive, decide and act with limited human intervention. This includes robots, drones and self-driving vehicles, as well as smart transport networks, cloud infrastructures and connected industrial environments.
The field brings together robotics, control, perception, computer vision, wireless communication, human-machine interaction, multi-agent systems, visualization and optimization.
Autonomous systems can contribute to safer, more sustainable and more efficient solutions by supporting demanding tasks, improving transport and logistics, strengthening public safety and enabling better use of resources in areas such as industry, forestry and agriculture.
Software
Software is the enabling technology behind AI, autonomous systems and many of today’s advanced digital infrastructures.
Software research in WASP covers methods and technologies for modelling, developing, verifying, deploying, maintaining and improving complex systems, including AI-based, machine learning-based and autonomous systems.
The research also includes software that uses autonomy, automation, learning or feedback, such as self-adaptive systems, self-repairing software, automatic programming and experiment-driven development.
Reliable software is essential for systems that people, companies and public institutions depend on. As software becomes more complex, research is needed to ensure that it remains secure, robust, efficient and understandable.