Digital Insect Monitoring: A Non-Lethal, Image-Based Approach
Monitoring changes in animal populations requires detailed data on species occurrence and abundance over time. For invertebrates, this is particularly challenging. Reliable assessments demand intensive sampling efforts, with repeated field visits to capture species that are active at different times of the year. This makes large-scale monitoring both logistically complex and costly.
Most traditional survey methods for invertebrates rely on trapping techniques such as Malaise traps, pitfall traps, or light traps. These approaches are typically lethal: specimens are collected, identified in the laboratory, and either discarded or preserved in collections. Maintaining such collections requires significant long-term resources and infrastructure.
Beyond logistical and financial constraints, the ecological impact of repeated and large-scale insect sampling remains insufficiently understood. This raises important ethical considerations, highlighting the need to minimize harm and to develop non-lethal alternatives wherever possible.
At the same time, advances in imaging technologies and artificial intelligence offer new opportunities to rethink how biodiversity data can be collected – enabling more efficient, scalable, and environmentally responsible monitoring approaches.
DiMON is funded by the Luxembourg National Research Fund (FNR) through the JUMP programme (PI: Lisette Cantú Salazar – Grant reference: 18303404). The project introduces a novel, non-lethal approach to monitoring invertebrates by combining automated imaging with advanced data processing.
At its core is an innovative device that captures high-resolution, multi-angle images of individual organisms as they pass through a trap system. The setup consists of a compact camera connected to a micro-computer and a mirror-based imaging chamber, allowing multiple perspectives of each specimen to be recorded simultaneously.
The device is integrated into existing entomological traps, guiding insects or other small invertebrates through the imaging chamber. As individuals move through the system, they are photographed in detail and then released unharmed back into their environment.
By capturing multiple views of each organism, the system significantly improves the accuracy of image-based species identification, particularly when combined with state-of-the-art deep learning models. This overcomes a key limitation of conventional single-image approaches and enhances taxonomic resolution.
The technology is designed to operate continuously and autonomously, generating large volumes of standardized data while reducing the need for frequent field visits and manual sample processing.
This project has the potential to transform how invertebrate biodiversity is monitored by offering a scalable, non-invasive alternative to traditional sampling methods.
By eliminating the need for lethal trapping, it supports more ethical research practices. At the same time, continuous automated data collection increases temporal resolution, enabling more precise tracking of population dynamics and seasonal patterns.
The system will be deployable across networks of monitoring sites, allowing data to be collected simultaneously over large spatial scales. When combined with automated image analysis, this opens new possibilities for real-time, large-scale biodiversity assessment.
The technology will also be accessible to a broader range of users, including citizen scientists, thereby supporting participatory approaches to environmental monitoring.
Ultimately, this innovation will contribute to more sustainable, efficient, and data-rich biodiversity monitoring systems – bridging the gap between technological advances and ecological research needs, and supporting informed conservation and policy decisions.
