Abstract: Crowd management is crucial to improve incidentresponse for enhancing Public Safety and National Security.
It is important tomanage crowd before, during and after the event. We deployed an applicationwhich collects GPS data using mobile phones with unique id of each user andafter processing, it gives an alert that whether anomaly is detected or not sothat concerned people and authorities can take actions accordingly.I.
INTRODUCTIONGathering of large masses of people atany place is normal situation but when this number increases or decreases expeditiously,then there is some anomaly, whether the reason for that anomaly is naturaldisaster (eg: Floods, Earthquake etc) or man-made disaster (eg: Terroristattacks, Protests etc).Recently tragedy struck Elphinstone roadrailway station stairs in central Mumbai during the morning rush hour on 29september 2017 resulted in death of 22people and nearly 40 injured terribly. One more casewas there in which people protested in favour of Ram Rahim Singh but after hisconviction, Situation like riots happened at various places, particularly atPunchkula and Sirsa where people gathered in very large masses resulted in deathof nearly 30 people and large number ofpeople were injured. So, We need crowd management or crowd control method forthe safety of people.Accordingto a forecast for global smart phone shipments from 2010 to 2017, more than 1.5billion phones are expected to be shipped worldwide. So, we can say thatmajority of people use smartphones.
Furthermore, smartphones have multisensingcapabilities like geolocation, light, movement, and audio and visual sensorswhich leads to a different and efficientways to opportunistically collect data and enabling new crowdsourcing applications. Crowdsourceddata is used to detect people. Crowd sourcing is of two types- Participatoryand Opportunistic. Participatory sensing means data is given by peoplevoluntarily as input and Opportunistic sensing means data can be generatedthrough sensors and automatically sent to server for computations. Smart phones already have several sensors:camera, microphone, GPS, accelerometer, light sensor, Bluetooth as proximitysensor. Thus, mobile people- centric sensing can be a scalable andcost-effective alternative.Geospatial data can be used to localize theindividuals. Geospatial data or GPS data is represented by numerical values forany object or individual in geographiccoordinate system.
In this work, we made an application tomake people remain safe by giving them alert about the mishappening occurrednear them. It also notifies the police and other related authorities about theevent so that they can take rescue steps as early as possible. In section II, we outline the related workon crowd management.
In section III, we describe application with which data iscollected and other details of dataset. Section IV describes implementation detailsof application at backend. In section V, we outline our results and withsection VI, we conclude our work.