Hu, DonghanBae, JosephLim, SolLee, Sang Won2023-11-022023-11-022023-10-29http://hdl.handle.net/10919/116591To mitigate the risk of chronic diseases caused by prolonged sitting, sit-stand desks are promoted as an effective intervention to foster healthy behaviors among knowledge workers by allowing periodic posture switching between sitting and standing. However, conventional systems either let users manually switch the mode, and some research visited automated notification systems with pre-set time intervals. While this regular notification can promote healthy behaviors, such notification can act as external interruptions that hinder individuals’ working productivity. Notably, knowledge workers are known to be reluctant to change their physical postures when concentrating. To address these issues, we propose considering work context based on their screen activities to encourage computer users to alternate their postures when it can minimize disruption, promoting healthy and productive behaviors. To that end, we are in the process of building a context-aware sit-stand desk that can promote healthy and productive behaviors. To that end, we have completed two modules: an application that monitors users’ computer’s ongoing activities and a sensor module that can measure the height of sit-stand desks for data collection. The collected data includes computer activities, measured desk height, and their willingness to switch to standing modes and will be used to build an LSTM prediction model to suggest optimal time points for posture changes, accompanied by appropriate desk height. In this work, we acknowledge previous relevant research, outline ongoing deployment efforts, and present our plan to validate the effectiveness of our approach via user studies.application/pdfenIn CopyrightArticle - Refereed2023-11-01The author(s)https://doi.org/10.1145/3586182.3616694