ExpRESAS: Expectation-Focused Requirements Elicitation for Self-Adaptive Systems
Research Project, 2025
– 2028
The goal of the project is to make self-adaptive software systems aware of their users’ expectations. A self-adaptive system can automatically adjust its behavior or structure. Many robots are self-adaptive. These systems can improve themselves, without requiring humans to manually reprogram or reconfigure them. Although humans should not be involved all the time, it is still important to consider humans´ needs and expectations. Expectations are assumptions of how a system might act. Current systems are not aware of humans’ expectations. In this project, I plan to help systems assess whether they meet humans’ expectations. If expectations are not met, the system can explain why and get input. Otherwise, it does not need to interact with the human and increase their cognitive load. In this project, I plan to develop a framework for self-adaptive systems to reason about human expectations, so that they can effectively understand the needs of their users and decide when to explain their behavior. In Year 1, I will develop a language to describe expectations using formal modelling methods. Afterwards, I will develop methods to understand the needs of human users and to explain a system’s actions. Finally, I will evaluate the framework using simulations and a real robot. My vision is to make systems intuitive and aligned with human needs. The project will help us to develop reliable, efficient, and responsible systems that can be understood and trusted by our society.
Participants
Rebekka Wohlrab (contact)
Chalmers, Computer Science and Engineering (Chalmers), Interaction Design and Software Engineering
Funding
Swedish Research Council (VR)
Project ID: 2024-05389
Funding Chalmers participation during 2025–2028