Date: September 24, 2025
Time: 10:00 AM (UK) / 09:00 AM (UTC)
Duration: 1 hour
Location: Online via Teams (free, registration required)
Capacity: Unlimited
Registration: https://events.teams.microsoft.com/event/3eda9a63-c338-419d-8bc9-0e4c41794184@7f025136-39f3-4f2c-a3c6-cbb1ce21b36a
Systematic site-specific monitoring of habitats in, for example, nature reserves, restoration sites or rewilding initiatives, remains a challenge, limiting our ability to objectively evaluate progress.
Our 1 hour webinar will explore using drone-based monitoring as a potential solution. We will give insights on how to achieve quality drone data collection under resource constraints. Using the example of mapping individual hawthorn shrubs in a rewilding site, we will explain the concept and power of U-nets (Deep learning) for habitat feature mapping and why they have so much potential. Finally, we will show how drone derived data can support allometry-based shrub biomass estimations.
Whether you work in biodiversity consulting, land-use planning, reserve management, research, or simply interested in habitat data collection and monitoring, this session will offer a useful example on how to approach drone-based monitoring.
Introduction (5 minutes)
Introduction to MAMBO and the webinar
· Introduction to MAMBO
· Why should we consider drone-based monitoring?
Drone campaigns (15 minutes)
What it takes to carry out a successful drone campaign
· The kit, the training, the protocols, the software
· Typical time and data storage needed to cover 1 km2
· Access to software and hardware
· The hurdles involved
· To sub-contract or not to subcontract?
Deep Learning (15 minutes)
Demystifying deep learning
· Explaining the concept and power of U-nets
· The training step: current challenges and solutions
· Testing and validating is important
· How code-savvy do I need to be?
· Access to software and hardware
Break (5 minutes)
From shrub map to shrub biomass (10 minutes)
An approach to deriving shrub biomass using drone data
· Introducing allometry with a Bayesian twist to reflect uncertainty
· Importance of destructive sampling
Q&A (10 minutes)
Participation is free but registration is required. Capacity is unlimited, and the Teams link will be sent to confirmed participants a few days before the event.
Register here: https://events.teams.microsoft.com/event/3eda9a63-c338-419d-8bc9-0e4c41794184@7f025136-39f3-4f2c-a3c6-cbb1ce21b36a