Reconstitution of the Journeys to Crime and Location of Their Origin in the Context of a Crime Series. A Raster Solution for a Real Case Study

In the region of Charleroi (Belgium), a series of criminal acts were
committed by the same group, using the same vehicle. The events were located in
space and time. The car used during these criminal activities was stolen (first event)
and was later retrieved (last event) after a period of 4 days of offences. Police
recorded a crucial clue: the total mileage covered by the vehicle between the first
and the last event was estimated with an admissible approximation. Thanks to this
information, we were able to choose the most probable journey-to-crime among
several scenarios. These depended on the combination of cost surfaces built with
distance propagation algorithms starting from each criminal event in raster mode.
The distance propagations were limited to the road network and the combinations
of the cost surfaces had to respect the chronology of the facts. The most plausible
scenario suggested that the criminals hided the car into a withdrawal site between
their activities. In order to improve the precision of the location of this withdrawal
site, we used a multi-criteria analysis taking account of the journey of the vehicle
and other environment variables. At the end of these treatments, the small stretch
of road that we isolated actually included the withdrawal site, as confirmed by the
police later