Looking at the natural disaster and man-made disasters happens does not seem to happen unannounced. There is always a clue to this occurrence either from the weather data or meteorological forecast and warning against some governmental policies. The focus of the solution is based on natural disasters. Natural disasters like Tremor, volcano eruption, Flood, Hurricanes herald by detection of molten magma under the rock, wind pressure originating from the aquatic splendor and typhoon from the middle east. These climatic or environmental conditions are processed and predicted by the relevant bodies of the specific geographical radius or territorial setup. The solution sends alert prior to impending dangers or disasters heralded by the climatic and environmental conditions. It also specifies the safety locations as the people evacuate from the disaster danger location neighborhood or specific radius from the danger geo spots. Using the Watson conversation service, the people holding the mobile application can also engage the Watson conversation bot to inquire about possible climatic condition mishaps looming.
The solution is divided into three parts:
1.The RESTFUL web service implemented as a suite of adapters in mobile first foundation services hosted on IBM cloud. The web services are convolution of weather data service, DB2 database, external web service endpoints from meteorological service as it concerns geo locations radius. These data sources are integrated using the machine learning associative rules to classify and detect likelihood of the classification that have led to disaster from the historical data.
The suites of web services also contain the CRUD (create read update delete) services implemented as an adapter for maintaining users on the platform.
2.Alert Notifications in email, SMS, push notifications to the mobile applications.
3.Watson Conversation web services bots. This is an intelligent agent that the user can interact with to garner information about a particular geographical area, destination in the case of tourists, and generally people living in natural disaster prone areas.
The user mobile application is implemented using the native application development because of the enhanced user experience and minimal dissipation of device battery. The mobile applications (android, iOS, Windows etc.) and web applications are the touch points of and the solution. The administrative layer of the solution would be administered by UN relief agencies or other relevant agencies to monitor, administer and action the broadcast of messages to the users having the mobile application.
The user downloads the app from the various application stores.
The geolocation , the Latitude and Longitude of the location is sought from the mobile phone by calling the specific resource form the operating system of the mobile phone (for instance Android GPS, iOS GPS). These data at every time is sought and send to the to web services in a synchronous manner and the result is presented for the user in humanly readable format. The Safety Location, Safety Report menus works as described. Other messages are generated and the backend or the command centers to furnish the user with relevant information based on the location, because the mobile application works by sending background processes to send the geolocation of users in disasters areas, at specific time interval. The background services get activated when the user approaches disaster prone areas. This is determined by calculation the distance form geolocation neighbourhood. This is calculated using the Manhattan distance and Euclidean Distance and the distance would be converted to Kilometers. The message broadcast by the background services informs the safety location and safety report messages that will be sent. Also the command center can also initiate a voice call to visually challenged. The solutions take cognizance of the respective disability of its users and the messages are marshalled in favorable channel.
The command center layer is built using Machine Learning computation in the Watson Studio.
The cognos service will be implemented to create
dashboard on the output data from the machine learning computation.
The dashboard will furnish the administrator/manager of the at the command centers.
The broadcast from the command centers will make use of REST web services calls to send push notifications, emails, phone calls, voice calls(using the Twillio third party services)