SEPAL: A Consistency-driven Programming Framework and Runtime Support for Human-Cyber-Physical Systems with Reliable Sensing and Dynamic Adaptation

In recent years, human-cyber-physical systems (HCPS) have become increasingly complex due to the widespread adoption of environmental sensing and behavioral adaption. Apart from the tight coupling between application logic and sensing-adaptation modules, such applications are mainly constrained by erroneous sensing and abnormal adaptation issues, often resulting in misjudgment of scenarios or adaptation behaviors that deviate from intended goals. Reliability in constructing and maintaining such application systems faces significant challenges, especially as human-cyber-physical scenarios exhibit dynamic uncertainties and evolving requirements, further exacerbating the development difficulty. To address these challenges, we designed and implemented SEPAL, a consistency-driven programming framework and runtime support for HCPSs with reliable environmental sensing and dynamic adaptation. SEPAL simplified the design of environmental sensing and behavioral adaption in HCPSs through a unified programming framework, and transparently managed the reliability of sensing and the unbiasedness of adaptation through its built-in two consistency-based services. SEPAL also provides a flexible browser-based management interface and a customizable interface design language for ease of usage. Case studies and evaluations have demonstrated SEPAL's facilitation of reliable support for various HCPSs, as well as the effectiveness and efficiency of environmental sensing and behavioral adaption capabilities.