WhatsApp: An Autonomous Alert System for Emergencies and Disasters

Abstract

The challenges derived from climate change, the presence of the El Niño phenomenon, and the intense rainfall that causes an increase in river flows underline the importance of having autonomous warning systems in emergency or disaster situations. This study focused on designing a conversational agent through the ManyChat platform, with the objective of alerting the population about the increase of the Piura River channel in Peru as the geographical area of study due to its overflow in March 2017. WhatsApp is configured as a fundamental component of this autonomous warning system since the proposed conversational agent is integrated with this platform, taking advantage of its wide penetration and accessibility in the community. The choice of ManyChat is supported by its ability to integrate with external services, such as Google Sheets and information access servers managed by technical-scientific entities. The use of technologies such as Google Apps Script and PHP made it possible to obtain and store data efficiently in the cloud, guaranteeing the availability and reliability of the information. The evaluation of the solution was carried out by means of a questionnaire addressed to 35 citizens of the localities of Piura and Castilla. The results indicate that 94.28% of the users are satisfied, and 91.42% highlight clarity and accessibility during conversational moments. For future research, the integration of a text messaging service is suggested to increase sociability, thus reinforcing the autonomy of the system in case of emergency or disaster.