WildINTEL

Context

Wildlife monitoring is critical to understanding, responding to and halting the current biodiversity crisis. Recent advances in biodiversity sensing, such as camera trapping, image classification technologies, citizen-science platforms, and machine learning, provide cost-effective wildlife monitoring.

However, there are still some bottlenecks, such as the high cost of manual image review and the lack of automated workflows. These constraints have hampered our ability to innovate and harmonise methods and tools for collecting and managing biodiversity monitoring data, and to take timely conservation and management actions.

Main objectives

The objective is to develop a cutting-edge coordinated wildlife monitoring system based on the Essential Biodiversity Variables (EBVs) framework. We will combine camera trapping, citizen science, artificial intelligence, and hierarchical models for the automated production of species population and community structure EBVs, with a precision that conventional monitoring schemes cannot match. This will enable stakeholders to obtain reliable and timely assessments of species conservation status and conservation actions to halt biodiversity loss.

The WildINTEL system will help mobilise and optimise the use of existing data and integrate camera-trap projects from other areas while supporting the analysis of the drivers of global change and biodiversity loss at spatio-temporal scales. We will focus on mammals, as they are condition sentinels as well as crucial indicators of ecosystem trophic integrity and global change.

Main activities

European biogeographical regions: Mediterranean, Continental, Alpine and Boreal. A protocol and application for the harmonisation of standardised image data collation will be developed. These data will be later uploaded to a global dataset infrastructure.

We will develop a transnational citizen science project in Zooniverse to enhance the general public’s involvement and knowledge of wildlife and to assist in image classification. These images will support the development of an automated artificial intelligence system for species identification and individual counts in biogeographical regions of Europe.

WildINTEL will produce automated spatio-temporal species and community EBVs data cubes in near real- time, corrected for imperfect species detectability and identification, thereby supporting efficient management decisions. EBV data will be periodically shared through the Global Biodiversity Information Facility.

Through several dissemination channels, including a website interface and several workshops, we will encourage new stakeholders to adopt the automated monitoring system and get involved in its upscaling. WildINTEL will implement a European stakeholder- led biodiversity monitoring, which will be maintained in the long-term.