Tools for automating image analysis for biodiversity monitoring – BBSRC

Your application should:

• outline how the tools proposed for development will improve the monitoring of biodiversity through automated analysis
• develop robust and verified software tools which can be broadly used across environmental science
• include research technical professionals such as research software engineers and data scientists

The full economic cost (FEC) of your project can be between £500,000 and £750,000. We will fund 100% of the FEC.

Scope

Biodiversity affects every aspect of life, from the air we breathe to regulating our climate and providing food and freshwater. We rely on it for our health and stability of natural resources like forests, aquatic environments, and grasslands. The health of the environment depends on the diversity of species that interact with it and vice-versa. Biodiversity monitoring is used to determine the stability of ecosystems.

The demand to improve biodiversity assessment is driving a need for more accurate, efficient, and timely data across a wide spectrum of spatial and temporal scales. Tools and technologies for imaging and the management of resulting data flows have huge potential to advance our biodiversity monitoring capabilities. These imaging technologies are enabling the collection of increasing amounts of complex biodiversity data across a broad range of parameters.

Imaging platforms can rapidly collect hundreds of thousands or millions of high-resolution images. However, the resulting large and complex datasets quickly exceed the capabilities of human analysts. There is a clear opportunity to improve use of automated image analysis to interpret a diversity of biodiversity data of increasing scale and complexity.

The aim of this funding opportunity is to support the development of software systems for image analysis in the laboratory, and in the field, to maximise the value of physical imaging technologies.

This investment will support the development of software toolkits, pipelines, and workflows for the automated analysis of images and videos. This could include, for example, automated image annotation and segmentation, high-throughput image analysis, the development and use of training data sets, and of automated pipelines.

Successful projects will develop robust and verified environmental image processing and software tools which can be broadly used across environmental science and can be maintained beyond this initial investment.

What we expect to see in applications

You will need to make the case for the capital investment required to develop tools, software, and learning datasets to address biodiversity challenges. It is not intended that significant investment will be made in new physical imaging technology, for example cameras.

It is not expected that this will primarily be a scientific research activity. The focus of this investment is on the development of software, pipelines, and systems, which can be used and maintained as tools for research. It will require dedicated effort from research technical professionals such as research software engineers and data scientists.

It is anticipated that approaches may learn from developments in other domains such as healthcare and image and vision computing, and that they will incorporate validation and verification methodologies building on environmental expertise.

We note that the work may include citizen science, or other approaches, to develop labelled training datasets, and will need to interface with the NERC Environmental Data Service or other appropriate data infrastructure.

Duration

The duration of this award is 36 months.

Projects must start by 29 February 2024.

Funding available

This is a capital funding opportunity, as such NERC will fund 100% of the full economic cost for equipment only. The full cost of your project can be between £500,000 and £750,000.

A total of £3.6 million funding is available

What we will fund

Costs can include staff time, but such costs must be directly associated to the creation, production, development, and preparation of an asset, so that it is capable of operating as a tool or system for the advancing our understanding of biodiversity and biodiversity loss. For a definition of an asset in the context of this opportunity please refer to the additional information section of this funding opportunity.

Development costs that are directly attributable to bringing the asset into working condition are allowable.

What we will not fund

• requests for resource, including staffing costs related to operation and maintenance of the tool (after the creation of the asset) and to user training
• single-user or single-project equipment
• development of novel technologies which are unlikely to generate a productive tool or asset
• applications have also been submitted to other award schemes during the same time frame (including all UKRI grant schemes)
• associated studentships

As this is a capital funding opportunity the following costs are also not permitted:

• pure research – original and planned investigation undertaken with the prospect of gaining new scientific or technical knowledge and understanding
• management overheads – unless that is performing technical development or solely purposed on the project
• incidental costs such as project office support, travel and subsistence, and training
• costs that only restore or maintain future economic benefit that was expected from the original asset

Find out more and apply by 18 October 2023 https://www.ukri.org/opportunity/tools-for-automating-image-analysis-for-biodiversity-monitoring