Context-Aware Chatbot Based on Cyber-Physical Sensing for Promoting Serendipitous Face-to-Face Communication

To promote the face-to-face communication reduced by COVID-19, we proposed and implemented a context-aware Slack chatbot based on Cyber-Physical sensing that helps colleagues meet more often in the same place. Our system periodically collects the user’s internal context through Slack (cyber sensing) and uses small BLE beacons distributed to colleagues and beacon scanners installed in a laboratory to sense physical attendance (physical sensing). In addition, the system notifies the user of recommended actions, such as lunch or coffee break, depending on the context determined by the Cyber-Physical sensors. We deployed the proposed system in a laboratory environment and conducted an initial experiment for six weeks. Experimental results confirmed that our system can encourage serendipitous face-to-face communication during periods when the frequency of attending school and going to work dropped due to COVID-19. It was also found that in an environment such as a laboratory, where a certain level of trust has already been established, the openness of the collected information can further motivate users to participate in the system.