An Intelligent cloud ecosystem for disaster response and management leveraging opportunistic IoT mesh networks

Access to emergency services like police, fire, rescue, and EMS is life or death during natural disasters. Disaster management faces three critical problems for emergency services - technology constraints (Network infrastructure), demand-supply management (a large number of victims to respond to, but limited on-field agents), information access (for on-field agents). In this paper, we present an end-to-end framework to enable reliable disaster response for emergency services. This framework solves the three problems described by introducing a unique system of collecting SOS messages from disaster victims, presenting and aggregating the messages to control center operators, and making this data alongside various offline tools available to on-field agents. The framework leverages a combination of Adhoc mobile networks based on widely used/readily available protocols and hardware to solve the technology constraints. We introduce a novel smartphone-based mesh network that leverages the radio modules already present in smartphones (BLE, Sound, Wi-Fi, Bluetooth) to complement custom hardware-based mesh networks (based on LoRa). SOS messages travel over multiple smartphones until they reach an internet-enabled device. On reaching the internet, we use contextual intelligence for determining the request context and helping emergency service agents prioritize and solve the request. We design an intelligent interface for control center agents to get an aggregated view on disaster victims and on-field agents, helping them make data-driven decisions to help the victims. The framework also provides the on-field agents with an interface to access data and communicate with the disaster victims, even in offline conditions leveraging the mesh network. The critical contribution of this paper is the framework's three-prong approach to support the victims, control center operators, and on-field agents. We present a walk-through for a pilot deployment of our framework alongside its qualitative and quantitative results and show how it can integrate with services like 911.