The objective of the current proposal is to develop an indigenous low-cost wearable pollution monitoring device that can be utilized for different applications. We aim to integrate the following innovations in our design – (i) the device will utilize edge computing to run ML at the edge to infer the pollution sources and its distribution, while suggesting suitable recommendations to its users based on their personalized needs depending on the age, medical history, application scenarios, and so on. (ii) Given the spread, lingering and contamination of pollutants at the spatio-temporal scales, particularly for the indoor environments, we plan to develop an architectural framework where nearby devices can coordinate with each other to provide a spatial pollution map to its users, both at the indoor and the outdoor environments. This involves considering the privacy, reliability and robustness of the data in decision making. Finally, we plan to consider factors like cost, performance and usability to develop the device and evaluate its success.
Accessibility Controls