**3. Empirical analysis**

*Natural Hazards - Impacts, Adjustments and Resilience*

shows the process of updating the index of missing people.

matched that may produce the index of missing people in the shelter. The updating of index notified the list of the missing people to all refuges in the shelter. **Figure 3**

The evacuation of the missing people needs to be confirmed by the shelter's authorities as shown in **Figure 4**. The administrative staff of MPCS will call the missing people and request to join the shelter. The local elected bodies have to contribute to bring the community people into the shelter. Mobile network operators will be given the mobile phone number and they have to send message and call for evacuation. Some people may go out and move to those shelters in emergency situations. Few people may visit their relatives and stay in the neighbouring shelters.

**44**

**Figure 4.**

**Figure 3.**

*Conceptual observation curve of the missing people in the shelter.*

*Updating the index of the missing people in the shelter.*

The smart devices are being used for the mobility detection of the car and people, but for the rescuing of people during cyclone that had very limited application. This research proposes a very simple method of detecting the missing people in the catchment area of a MPCS by using Wi-Fi scanner in order to rescue death tolls before a cyclone landfalls. The proposed method includes indexing of the community people in the catchment area, indexing of people in the shelter, determining the missing people, and updating the index of missing people.

This study used smartphones and a Wi-Fi scanner. A MacBook Pro laptop was used as a Wi-Fi scanner. Experiment was conducted on 11/12/2020, 8/01/2020 and 9/01/2020. The data obtained on 9th January 2020 was used for this analysis. Total 90 participants joined in the experiment. The experiment was a control experiment to simulate the field conditions. All participants were requested to turn off their hotspot of smartphone. Few participants were asked to turn-on the hotspot of their smartphones and Wi-Fi scanner detected the MAC address with SSID of the participants. This process was continued till all participants were detected with a time interval of five minutes. The penetration rate was 100% for this experiment that means all participants were detected during the experiment. The MACs and SSIDs are not disclosed for the privacy reason. Only number of people detected and participated are being used for this study.
