Wenying Chang The thesis aims to investigate how the Sleep Robot system (a Sleep Robot, its application, and servers) can adopt a data-enabled loop to make optimal use. The research has three main objectives: (1)understanding the needs and concerns of using the Sleep Robot system in different user profiles, (2)building a data loop of the Sleep Robot system with subjective and objective data, and (3) identifying the design principles of long-term engagement in a data loop.To answer the research question and fulfil the objectives, the thesis adopted a four-diamond design process. In the discovery phase, the research reviewed the literature on personal informatics and sleep solutions, then plotted the Somnox data ecosystem. In the refining phase, the user research was performed to obtain profound insights, create a journey map, and user profiles by using a sensitizing diary, fetching data from servers, creating data visualization, and conducting interviews. The user research found that users stick to standardized breathing settings without engaging themselves in self-experimenting with breathing settings due to the lack of guidance and support. Thus, the project’s design goal is to let users feel guided and at ease during self-experimenting different breathing settings.By going through creative sessions and user testing in the developing phase, the research found out four essential design elements to achieve design goal: (1) step by step learning experience, (2) dynamics of breathing rate, (3) growing together through empowering the breathing recommendation system, and (4) breathing with a specific goal which reflects causes of poor sleep. Eventually, the proposed design integrated these four elements into three levels of personalized breathing rhythms, including customized breathing, optimized breathing, and contextualized breathing.In the evaluation phase, the proposed design was evaluated through user scenarios and high-fidelity prototypes. The results showed that users highly appreciated the guidance and ease through measuring personal conditions to tailor three-level breathing rhythms. This study emphasizes the transition role of the proposed concept from three-level of personalized breathing with a data loop and, ultimately, toward the adaptively synthesized breathing.

Master Thesis: TU Delft repository