Knowledge Base
Read articles written by our experts on a range of topics drawn from across the project
The TAiM project: Outcomes and insights
As the project comes to a close, we reflect on what we have learned and where we will be going next.
Open PDFQuickBeam: ground-truthing for AI mapping trust
A new mobile app for mapping data validation was developed in the project, in this article we explain why it was developed and how it works.
Open PDFTranslating field surveys to satellite pixels
Understanding how best to comparing high resolution field data against coarse satellite pixels is a challenge for map validation. In this aricle we explore the problem and propose some solutions.
Open PDFUnderstanding social trust in AI mapping
An exploration of the simple criteria and steps can be used to self-assess social trust in AI systems and which can be applied to AI derived maps.
Open PDFAn overview of AI algorithms
A basic primer on AI algorithms and how they relate to the machine learning (ML) algorithms commonly used in mapping projects.
Open PDFIntroduction to explainainable AI (XAI)
What is explainable AI, and how you can use it to bring greater understanding to how an AI model is making decisions.
Open PDFAlgorithm spotlight: Random forest
A deep dive into one of the most widely used AI mapping algorithms.
Open PDFAlgorithm spotlight: Neural networks
What are neural networks and deep learning, and how do they work?
Open PDFAlgorithm spotlight: Convolutional neural networks.
An overview of how neural networks have been extended to make sense of image data.
Open PDFAgrimetrics Agricultural, Woodland, and Landcover algorithms
Overview of new mapping products developed by Agrimetrics during the project that utilises their novel convolutional neural network approach.
Open PDFMapping peat condition
How have the James Hutton Institute used machine learning to map peatland condition across Scotland.
Open PDFMapping soil carbon
How have the James Hutton Institute used machine learning to map soil carbon across Scotland.
Open PDFSurveys of grassland composition and condition
Grasslands are a rich source of biodiversity and can be significant carbon stores. Learn how Highlands Rewilding have been surveying the composition and condition of these important ecosystems.
Open PDFMapping, monitoring, and muddy boots
A trip to Tayvallich Estate to survey vital saltmarsh habitats.
Open PDFSurveying peatland in the field
Highlands Rewilding shares their story on surveying peatland depth and condition.
Open PDFWoodland surveys in the field: Making sense of the woodland census
How have Highlands Rewilding collected accurate tree census surevys, and why is this important for habitat restoration?
Open PDFAn Introduction to Our Drone Surveys
An overview of the drone lidar surveys that The James Hutton Institute have collected to map our study sites.
Open PDFElevating Insights: Harnessing Drones for Precision Data Collection
A technical description of how The James Hutton Institute utilised drone-based data capture and processing for the TaiM project.
Open PDF
The TAiM project: Outcomes and insights
As the project comes to a close, we reflect on what we have learned and where we will be going next.
QuickBeam: ground-truthing for AI mapping trust
A new mobile app for mapping data validation was developed in the project, in this article we explain why it was developed and how it works.
Translating field surveys to satellite pixels
Understanding how best to comparing high resolution field data against coarse satellite pixels is a challenge for map validation. In this aricle we explore the problem and propose some solutions.
Understanding social trust in AI mapping
An exploration of the simple criteria and steps can be used to self-assess social trust in AI systems and which can be applied to AI derived maps.
An overview of AI algorithms
A basic primer on AI algorithms and how they relate to the machine learning (ML) algorithms commonly used in mapping projects.
Introduction to explainainable AI (XAI)
What is explainable AI, and how you can use it to bring greater understanding to how an AI model is making decisions.
Algorithm spotlight: Random forest
A deep dive into one of the most widely used AI mapping algorithms.
Algorithm spotlight: Neural networks
What are neural networks and deep learning, and how do they work?
Algorithm spotlight: Convolutional neural networks.
An overview of how neural networks have been extended to make sense of image data.
Agrimetrics Agricultural, Woodland, and Landcover algorithms
Overview of new mapping products developed by Agrimetrics during the project that utilises their novel convolutional neural network approach.
Mapping peat condition
How have the James Hutton Institute used machine learning to map peatland condition across Scotland.
Mapping soil carbon
How have the James Hutton Institute used machine learning to map soil carbon across Scotland.
Surveys of grassland composition and condition
Grasslands are a rich source of biodiversity and can be significant carbon stores. Learn how Highlands Rewilding have been surveying the composition and condition of these important ecosystems.
Mapping, monitoring, and muddy boots
A trip to Tayvallich Estate to survey vital saltmarsh habitats.
Surveying peatland in the field
Highlands Rewilding shares their story on surveying peatland depth and condition.
Woodland surveys in the field: Making sense of the woodland census
How have Highlands Rewilding collected accurate tree census surevys, and why is this important for habitat restoration?
An Introduction to Our Drone Surveys
An overview of the drone lidar surveys that The James Hutton Institute have collected to map our study sites.
Elevating Insights: Harnessing Drones for Precision Data Collection
A technical description of how The James Hutton Institute utilised drone-based data capture and processing for the TaiM project.