Ulo Sare
CEO, Tartu Smart City Lab
Estonia

Biography
Mr. Ülo Säre was born on 13th june 1974 in Tartu, Estonia. Ülo obtained his BSc in Tartu University where he studied mathematics and economics and started his first IT project at age of 17. During last 20 years he established couple of IT companies and was member of Estonian Association of Information Technology and Telecommunications. Since 2013 he is working as the CEO of technology company Reach-U, which is biggest European Location Based Solutions provider. Reach-U core competence is in mobile positioning from the Mobile Operators core network and experience of building solutions for National Security, Public Safety, Civil Warning and Transportation. Reach-U is also active member and one of establishers of Tartu Smart City Lab - the cluster of smart e- and m-city solutions is designed to create an innovative environment in Tartu by bringing together businesses, citizens, public authorities, R&D institutes and structures that support innovation.

Abstract
Actionable Transportation Analysis by turning all our GSM-phones into sensors

Mobile radio access network has lots of location data. With existing technology it is possible to use all the mobile-phone owners as censors. Both citywide and nationwide. It is possible to detect typical routes of people, traffic congestions, spatial distribution of homes and workplaces etc. The purpose of the presentation to give an overview of new enables in technology for planning the infrastructure and transportation as well to share experiences and tips on how to achieve a successful outcome. R-U/DG has performed analysis on population movement patterns on daily, weekly and seasonal timescale:

1) Origin-destination (O/D) matrix estimation – the whole territory is divided into “n” smaller areas, and one estimates the number of people moving between any pair of these areas.
2)Traffic density estimations
3) Estimation of proportions people who use different means of transportation, e.g. metro vs. road traffic.
4)Estimation of travel speed between areas, depending on time of day, day of week, season.
5) Estimation of location semantics within trajectories of individual homes, work locations, etc. to be used in aggregated statistics.
6) Estimation of population density and its variation
7) Difference between local people and foreigners (roamers)
8) Combining available additional info with positioning-based info. E.g. compare behavior of different age groups or genders.

Pitfalls to consider when implementing location analysis:
1)The passive positioning data has spatial uncertainty levels significantly higher compared to GPS measurements and it is not trivial to get reliable estimates of interested variables.
2) The amount of mobile data is very large – typical volume is several billion events per day. Specific software solutions are needed to handle this data volume in reasonable time, on some use-cases in real-time.
3) Its easy to lose in all the possibilities. The “actionable” check is good tool for validating different options and analyses.