EDA-sense: Dynamic Feedback Control of Sympathetic Arousal

Abstract In our society, individuals experience stress on a daily basis. This condition is prevalent and is challenging daily lives. One way to mitigate stress impact is to use concurrent information related to underlying cognitive stress to act on the environmental stimuli leading to it. In our work, we harness electrodermal activity (EDA) collected by wearable devices as a proxy for neural activation to develop a real-time control loop for stress pacification. We propose a model of EDA response to facilitate simulation design and testing. We find the hidden cognitive state/stress level, detect the distress regions and apply continuous control by updating the control-signal with a on-off controller. Finally, we demonstrate the validity of our method on both simulated and experimental real-time data collected by exposing subjects to different stress drivers. In conclusion, the proposed work suggests development in automatic arousal control, by constructing a human machine interface that could make the subject aware of the physiological state and provide feedback on what actions to take to achieve certain desired neurological status.