The ubiquitous computing gives unlimited possibilities to create intelligent solutions that increase efficiency in many areas of life. For example, the location of smartphones informs where and when the user was located, which allows you to simulate scenarios of the spread of infectious diseases, develop efficient contingency plans and quickly anticipate the emerging congestion. These solutions, although groundbreaking, often meet with criticism related to the protection of personal data privacy. Privacy protection and legal regulations prevent the widespread use of such data in creating innovative solutions.

            Privacy protection at the expense of data potential or vice versa, sacrificing people’s privacy for the purpose of creating new solutions will soon be no longer a problem. The solution for that was recently presented by scientists from the Wrocław University of Life Sciences in cooperation and the University of Auckland in their latest article published in the Computers, Environment and Urban Systems journal. The presented solution provides full-potential data on the location of the population, which can be used in almost any field related to human mobility, while protecting the privacy of the population. The idea is based on the generation of artificial movement trajectories whose characteristics are similar to the original data used to calibrate the data generating model. In this way, artificial trajectories do not coincide with the real movement of the population, but provide the same information.

            The developed model, called WHO-WHERE-WHEN (3W) is at the same time a human mobility model. During the calibration phase it creates an image of the mobility of a given area, so that it is also possible to generate hypothetical scenarios. For example, it is possible to assess the impact of adding the new residential area in the suburbs of large urban centers on the formation of congestion. Compared to the best solutions in this field, the 3W model has obtained 35% better accuracy in replicating movement trajectory characteristics while increasing the flexibility of the solution and the range of reflected population mobility characteristics.

Fig 1. Map showing a comparison of population density at a selected moment in the New York State area, calculated from generated data (left panel) and real data (right panel). Author of the map: Barbara Kasieczka.

            The conflict between the Internet of Things and privacy is the greatest constraint on the further development of many areas. The 3W model is expected to solve this problem in the future by creating universal access to location data, thus opening up a new market for mobility-based services, where the possibility of participation will not be limited by the unavailability of data, and equalizing the opportunities for corporations, small and medium-sized companies and academic entities.

            Article “Population mobility modelling for mobility data simulation”, authors: Kamil Smolak, Witold Rohm, Krzysztof Knop and Katarzyna Siła-Nowicka, is available at DOI:10.1016/j.compenvurbsys.2020.101526.