AI-based approach to foster access and scale to real-time ground forest analytics

Traditional forest measurement methods are labour-intensive, costly, and prone to errors, hindering scalability and accessibility. While remote sensing offers valuable insights, it falls short in capturing crucial details beneath forest canopies, leading to inaccuracies in carbon stock calculations. This study introduces a citizen-science-based approach that leverages smartphone technology and artificial intelligence (AI) to democratize and enhance real-time forest analytics.

The methodology employs a mobile application that guides users through Point Sampling, as described by Bitterlich (1948), eliminating the need for specialized tools and expertise. Users capture geotagged photos at designated points within the forest, which are then analysed by a computer vision model to reproduce forestry equipment like a prism, counting tree trunks, identifying their species, and determining ground characteristics, paired with remote sensing inputs. By integrating smartphone capabilities with AI-driven analysis, the platform enables rapid estimation of forest parameters, including basal area, biomass, vegetation structure, and biodiversity insights.

The qualitative results highlight the efficacy of this approach in overcoming the limitations of traditional field forest inventory methods. The user-friendly interface of the mobile app empowers local communities to actively participate in data collection alongside experts, fostering inclusivity and environmental stewardship. This innovative approach not only reduces costs and time associated with forest assessments but also promotes community engagement and contributes to more sustainable forest management practices.