Towards Forest Dynamics' Systematic Knowledge: Concept Study of a Multi-Sensor Visually-Tracked Rover Including a New Insect Radar for High-Accuracy Robotic Monitoring

Forest research is essential for understanding the global carbon cycle and multi-scale forest
decision-making, management, and conservation. The need for systematic forest monitoring
approaches can be addressed using recent achievements in robotics and computing. Compact
sensors and associated data management techniques are revolutionizing this topic. We propose a
compact autonomous station hosting a low-range radar for high-detail nocturnal insect (and small
species) surveillance with the potential being extended to other niches of stationary data collection
aiming mainly at forest animal activity but potentially including vegetation and tree phenology and
forest floor conditions. Moreover, we introduce a novel multi-sensor visually-tracked rover concept
allowing complex forest phenology monitoring with multiple potential applications. The proposed
idea of rover automatic tracking with total station allows for collecting data with centimeter
accuracy. A deployed dense control network facilitates it. Furthermore, the network is utilized
for accurate tree measurement to interpret the collected sensor data. The demonstrated sample
data confirm the effectiveness and high potential of the proposed solutions aiming at systematic
forest dynamics monitoring.