Kido | our solution
page-template,page-template-full_width,page-template-full_width-php,page,page-id-120,ajax_fade,page_not_loaded,,qode-title-hidden,qode_grid_1300,footer_responsive_adv,hide_top_bar_on_mobile_header,qode-content-sidebar-responsive,qode-theme-ver-16.6,qode-theme-bridge,disabled_footer_bottom,wpb-js-composer js-comp-ver-5.5.1,vc_responsive


Starting from mobile big data

Mobile operators are collecting, every minute, aggregated information of millions of anonymous users in their network. This data, which includes metadata such as time stamp and position of billions of telecommunications events, represents an accurate picture of the collective human activity.

By means of big data and social physics, our private-by-design, patent pending algorithms analyze these massive data unveiling the underlaying mobility pattern of the population at the same time that privacy and anonymity of every user is preserved.


Problems to solve: fragmentation and accuracy

The raw data obtained from the network is unstructured and noisy. This data needs to be processed and filtered to select only those events that contribute to the definition of the human mobility patterns.

In addition, we as users do not use continuously our cell phone, thus gaps and blind spots (from minutes to hours) are very common, while network is not homogeneous, which makes it difficult to accurately identify patterns precisely at roads or streets.

These problems have represented a big barrier in the design of new methods for leveraging this data.

Social physics to eliminate fragmentation

In the last decade, a promising and emergent discipline was born in the interdisciplinary confluence of physics, sociology and big data: social physics. Computational and statistical methods borrowed from physics are used now to describe and predict the behavior of human crowds.

Following the innovations of this new discipline, we have designed proprietary, patent pending algorithms to structure mobility data and reconstruct the missing information with state-of-the-art machine learning and predictive methods inspired in computational quantum physics.


Machine learning to enhance accuracy

With this information, our trained algorithms run millions of routes agent-based simulations, selecting those components that better fit the observed events, determining the most likely path followed by the crowd.

As a final result, we integrate all meaningful insights derived from analyzing crowd mobility patterns: origin and destination, routes, average speed, trip distance, stops, means of transportation and seasonality of behavior.

delivery formats

The insights generated by our technology can be tailored to adapt smoothly to your internal processes and systems, so you can quickly integrate the knowledge generated by our solutions to your decision-making routine.

detailed reports

Development of deep analysis focused on answering specific questions that your organization needs to solve in order to take a straightforward decision.

interactive maps

Information displayed on flexible tools with the possibility of interaction and generation of dynamic analysis so you can explore and generate your findings a long the way.

importable data

Data and metadata generated in several standard formats (e.g. csv, shp, GeoJSON) to be imported by your geospatial system, with the desired level of granularity.