- Browse/Search the Program
Search
Program Calendar
Browse By Day
Browse By Room
- Information Menu
Search Tips
Virtual Exhibit Hall
- Navigation and Settings Menu
Personal Schedule
Sign In
Exploring AI perspectives in crime deterrence and law enforcement: an application on illegal timber logging in Romania
Sat, September 6, 8:00 to 9:15am, Communications Building (CN), CN 3104Abstract
In February 2020, an infringement procedure was launched against Romania for illegal logging and for failure to properly implement the EU Timber Regulation. Since then, Romania was putting a lot of effort in developing an electronic timber traceability system (SUMAL) assorted with administrative and penal indictments for illegal logging. The SUMAL 2.0 version released in 2021 is collecting for each timber transportation information about the vehicle, timber load and destination, documented with pictures of each timber load and a GPS recorded trajectory. The tracking system generates a huge data basis, every year cumulating some 5 million new entries.
Taking into account the pace of data collection and the legal prescriptions delay, increasing the capacity to process the SUMAL 2.0 recorded data is a critical factor in sanctioning the illegal timber harvesting and trade. In this respect, new perspectives are brought by the possibility to use AI in data processing for forest crime deterrence, for example in identifying the upload of fake pictures. This study presents the results of the experimental use of Gemini soft for a period of three months in Romania that helped to process 334 thousand pictures and to identify around 10% of them with non-conformities and to detect as well around 7% problematic transportation cases. Beside the advantage to reduce by 90% the human-required detection effort, one of the most important advantages of using AI-based detection in timber traceability system stays in the possibility to perform risk analysis per SUMAL 2.0 user or per geographical region.
As conclusion, the enforcement of the forest law may have a very useful and efficient tool in AI-based applications to detect irregularities. Even if irregularities may not be a penal offence per se, they may however indicate an increased risk for trespassing behavior that will be further confirmed through regular controls.