Using passive anonymous mobile positioning data & aggregation

Using passive anonymous mobile positioning data & aggregation analytics to enhance toolsets for flood relief agencies

Mohammad Nasim

CSE, Lingayas Vidyapeeth

Faridabad, India

[email protected]

GV Ramaraju

Lingayas Vidyapeeth

Faridabad, India

[email protected]

Abstract— after a natural flood disaster, relief agencies need to know where influenced individuals are located, what things are required, and who is protected. This data is incredibly tough and often difficult to get through traditional data gathering strategies [1] in an opportune way. This study can give bits of knowledge in near-real time to help relief agencies to facilitate their work and fill pivotal holes in information amid disasters.

Utilizing mobile positioning data and aggregation data analytics we can design a better solution for relief agencies with the goal that they could plan the activities better. This study explains about the data analytics of Location, displacement and population density with comparison to historical data of few weeks that can predict good quality results and tools to handle the response phase of flood disaster.

This investigation clarifies the advancement of Cellular Network Coverage, Power, Displacement, and Safety check diagrams based upon data aggregation analytics on Mobile positioning data of the cellular phone we carry. This study encourages agencies to address the critical gap in information they frequently face when reacting to cataclysmic events. Study graphs give data about where populaces are found, how they are travelling, and where they are feeling themselves safe amid a cataclysmic event. We can impart these accumulated bits of knowledge to relief agencies so that they have the data they need to help communities recuperate and revamp if debacle strikes.

Get quality help now

Proficient in: Data

5 (339)

“ KarrieWrites did such a phenomenal job on this assignment! He completed it prior to its deadline and was thorough and informative. ”

+84 relevant experts are online
Hire writer

Likewise this will also help to assist them in satisfying the necessities of migrating, food, shelter, electricity, water and other fundamental needs of the people in question.

Index Terms— MPD, data-aggregation, flood rescue, Mobile computing, Global Positioning System, Cellular networks


In the past decade much research has focused on Mobile location data either through any mobile app or mobile GPS data with a condition location ‘on’. In realty, at the time of disaster, our high expectations from the victims that they will use their mobile with location ‘on’ continuously and daily in the response seems unpractical. This study proposes passive mobile positioning data [2] managed by mobile network operators as a base to develop crisis tools. It remains unclear that why this MPD is still in the hands of very few government agencies. Answer can be the data privacy policy of any country and security is one of the reasons to hide this from public. This study proposes here to mine the data without hitting the security, safety and privacy policy. This study explains the use this MPD in various GIS related platform to solve many real world problems.

Amid flood debacle, where here has a fiasco damaged cellular network? Where can individuals get to a cellular system to connect with their loved ones? Where are individuals ready to charge their gadgets in a fiasco influenced region? What does this infer about power accessibility? What numbers of individuals have been dislodged from their home city after a catastrophe? Which towns are these individuals dislodged from? Which towns are dislodged individuals moving to?


To move more nearer to the issues of people and relief agencies at the time of any disaster, we have taken few case studies mentined below to analyze different aspects during disaster. All the below case studies has been taken from different crisis and data has been aggregated using one popular mobile application with GPS location ‘on’.

A. Contextual analysis: Fort McMurray wildfire

On May 3, 2016 flame tore through Fort McMurray as 88,000 individuals fled for their lives. We see that before the emergency started (left plot) individuals were situated in the town in anticipated numbers, yet immediately cleared throughout the following 24 hour time frame (map turning red) as shown in Fig 1 below. Indeed, even inside only an hour and half into the emergency we can tell that people are emptying the town (the red shading demonstrating low estimations of individuals present contrasted with pattern information). This sign turns out to be considerably increasingly clear and steady as the emergency advances.

Fig. 1. Populace during McMurray Fire (May 3, 2016)

Populace here alludes to the number of inhabitants of mobile phone users. These consolidated maps can give an intermediary to populace density and their travelling previous, amid and after catastrophes.

In the above picture, the blue line that extends slantingly over the map is Highway number 63, which was the basic evacuation path for maximum people in McMurray. Notice that the blue line turns out to be significantly denser somewhere in the range of 1 and 3 A.M. nearby time [3]. Reports from the prevailing press distributed that evening uncovered that numerous drivers wound up having to “camp” along the expressway whole night. Red pixel indicates they are abandoning out from the terminated district to some more secure spots.

Contextual analysis: Kaikoura Earthquake in New Zealand

Take the map shown below of the Kaikoura Earthquake in New Zealand as another model. The debacle maps for the seismic tremor demonstrate the position and displacement of individuals in Kaikoura following the calamity. One day after the seismic tremor, we see that the populace starts to move out from the city as shown in Fig 2 below. Utilizing news feeds, we can cross approve that occupants of Kaikoura were moved to Christchurch, 200 kilometers away. A few days after the fact, we see from the fiasco maps that people are coming back to Kaikoura, apparently to fix and reconstruct their locale. [4]

Fig. 2. Kaikoura Earthquake and Populace representation of 3 days

B. For Cellular(Mobile) Network Coverage Survey

Contextual analysis: Volcanic Eruption in Guatemala

Fig. 3. 2G network before Volcanic Eruption – Guatemala

Heat Map, which demonstrates the 2G network map in Alotenango, Guatemala on April 22, 2018 before the volcanic emission as shown in Fig 3 above. The perception below shows 2G network diagram on June 10, 2018, 4 days after the volcanic emission as shown in Fig 4 below. In light of the sharp increments in red and orange regions, also full blank spots on the map, we can see a sharp drop in network use in the territories encompassing the site of the ejection.

Fig. 4. 2G network after Volcanic Eruption – Guatemala

C. Battery/Power Charging Survey in Affected areas

Contextual analysis: Storms in Ranchi, India

The figure beneath shows changes in number of mobile phones backed with power, communicated as the baseline average, in zones close to Ranchi, India on May 28, 2018 as shown in Fig 5 below, which experienced power blackouts in the wake of lightning and thunder storms. While control accessibility was not influenced in the downtown area of Ranchi where the zone is as yet light blue, there was a sharp decline in the quantity of mobile phones interfacing with power in territories outside the city, featured in red. [5]

Fig. 5. Mobiles communicated in zones close to Ranchi, India (May 28, 2018)

D. For polulation Displacement Survey:

Contextual analysis: Wildfires in Northern California

In October 2017, obliterating rapidly spreading fires devastated an expected 8,900 structures in the Napa Valley area, constraining 100,000 individuals to move, huge numbers of who remained uprooted from their homes well after the flames were smothered. The figure underneath demonstrates the net difference in populace amid this time, revealing the absolute change in the quantity of individuals in every city in the week of December 4, 2017, around 2 months after the flames began, contrasted with the week prior to the flames began as shown in Fig 6 below.

Fig. 6. Populace representation (Wildfires- Northern California – 2017)

The points in purple, blue, and green show urban areas with a decline in individuals, while the orange and red points show urban communities with an expansion in populace. From this representation, we can see that about the majority of the urban areas in the Napa Region demonstrate some diminishing in populace after the flames, including Santa Rosa, Rohnert Park, Napa, and Vallejo [7] . We can likewise observe the urban communities where uprooted individuals went in huge numbers, including real urban areas outside the risk zone, for example, San Francisco and San Jose.

E. For Safety Check Survey:

Contextual analysis: Safety Check

Fig. 7. Fort McMurray Safety Check Indicators

Within 24 hours of actuating Safety Check, we see that there are far less individuals than expected in the town of Fort McMurray. Regions that are shading coded red reflect much lower quantities of Mobile people there contrasted with a similar time the prior week as shown in Fig 7 above. This bodes well since these areas are influenced by the rapidly spreading fires and have in this way been evacuated.


Study is being done by taking the report ‘Feasibility Study on the Use of Mobile Positioning Data for Tourism Statistics’ [4] as a base of the research and then refining the same technology and processing of data to do various spatial analytics specific to flood disaster.

The procedure contains the accompanying areas: the extra readiness of occasion information, frame creation, data aggregation and estimation. The base data filtered and arranged by MNOs depends on system events that determine a particular customer’s presence in reality. Extra readiness may incorporate geological referencing, the removal of non-human operated cell phones, checking the time and territory coverage of the data, managing missing qualities, and so forth as shown in Fig 8 below.

Fig. 8. Mobile positioning data processing steps

At the point when individuals utilizes a cellphone , their x and y coordinates are received at regular intervals and gets recorded as passive MPD data by MNO’s. This Location data is utilized in various ways, such as delivering a feature or content that is most important to individuals. For instance, it enables us to send notifications to individuals’ news feeds in focused pursuit regions after a kid has been stole, or Safety Check warnings to those in areas influenced by catastrophic events.

Same geolocation information, when amassed and de-recognized, gives important data to help relief agencies after a catastrophic event. Conglomeration not only helps to maintain security, yet additionally makes the information increasingly usable and interpretable to relief agencies by isolating disturbance from clamor, and in this way diminishing the intermediate compilation steps required to move from data to insights for action.

The debacle maps datasets are accumulated crosswise over time in the accompanying ways:

• Temporal accumulation: While data is required over time frequently amid a debacle. Consequently, we share information at consistent time lapse (e.g., Every 24 hours, every six hours, and every hour).

• Spatial conglomeration: We totaled geo-localized spots to 360,000 square meter tiles or nearby administrative limits.

• Spatial smoothing: Once we have determined every metric (e.g., the count of individuals in administrative or pixel unit x amid timespan y), spatial smoothing is build. For each spatial area, we process a weighted average of the value in the tile itself with the values in adjacent square tiles; tiles that are nearer have a greater commitment to the final outcome . This neighborhood averaging results in a map with a smoother, more clear signal, lessening commotion because of irregular variation while safeguarding the key signal and further securing privacy.


Utilizing the information and total procedures depicted above, we can create three exceptional datasets:

• Population Diagram: Metrics showing the density of the MPD populace in each square tile.

• Movement Diagram: Metrics identified with populace displacement between tile sets.

• Safety Check Diagram: Metrics showing the density of Safety Check registration versus all solicitations for each tile.

A. Polulation Density Diagrams

By totaling geolocation information, we can demonstrate a smoothed portrayal of what number of individuals with area administrations empowered is utilizing Cell telephone in each regulatory district or guide network for each timespan.

Cite this page

Using passive anonymous mobile positioning data & aggregation. (2019, Nov 26). Retrieved from

Let’s chat?  We're online 24/7