This use case is from South London Partnership – InnOvaTe “IoT” Project.
Outcome
Using sensors to monitor wildlife corridor conservation sites to help identify biodiversity changes within Richmond and Kingston to inform how the council manages council-owned land.
Sectors
Local authority, Green and climate change teams
Key Stakeholders
BioDiversity Team, Institute of Zoology – Zoological Society of London (ZSL), Data Team, Environment team
Overview
The South London Partnership (SLP) designed an Internet of Things (IoT) trial to capture wildlife data in conservation sites to help identify biodiversity changes within Richmond and Kingston. Utilising the remote sensing methods, officers investigated species richness and activity on key wildlife corridors. A video about the study was posted on Kingston Council’s YouTube account.
About the Author
The InnOvaTe Programme is an Internet of Things initiative by South London Partnership (SLP) to “pilot and research” IoT across the 5 London boroughs of Croydon, Merton, Richmond upon Thames, Sutton, and the Royal Borough of Kingston upon Thames. The programme looks at ways to generate economic growth, support local businesses, help people live better, healthier lives and assist with addressing the climate emergency. The project assessed 150 IoT ideas for the councils concerned, implementing 48 of them successfully over 18 months. The programme was formally completed in March 2023.
Case Study Challenge
In 2021, Richmond and Kingston Councils embarked on a mission to revitalise their approach to nature recovery, with a particular focus on the Sites of Importance for Nature Conservation (SINC). This initiative included a comprehensive resurvey of the flora and habitats within these sites, a process not undertaken since the early 1990s. The goal was to integrate this new data with additional information to create a strategy that enhanced site connectivity, bolstered climate change resilience, and prioritised actions for habitats and species. To support this strategy, the councils considered the implementation of IoT connected cellular camera traps. These advanced devices offer significant advantages over traditional non-connected technology by providing real-time data for immediate analysis and response. Strategically placed in various quiet locations across the borough, these camera traps capture static images triggered by movement. The images are then uploaded to the cloud, enabling analysis by officers, community groups, volunteer citizen scientists, and the Institute of Zoology using machine learning technology.
Use Case Design Objectives
This trial is part of the councils’ efforts to update biodiversity information with more accurate and current data on local species. Traditionally, data was collected manually through surveys. The new data will be used to improve conservation efforts and address gaps in knowledge about the functionality and health of the boroughs’ wildlife corridors.
The use of sensing camera technology will enhance research methods by allowing the assessment of wildlife presence and health. Stationary devices placed throughout habitats enable officers to observe cryptic species and their behaviours without direct human interference, which can alter animal activity. Additionally, the capability to remotely send data reduces human effort and facilitates extensive investigations over larger geographical and temporal scales than previously possible.
The primary aim of this use case is to confirm the presence of target species, which will inform how the council manages council-owned land. For connecting lands between sites that are not council-owned, the council can advocate for the animals with landowners or offer to manage the land with the owner’s permission. Furthermore, if the data allows, officers aim to determine the population status of each species and identify which sites are most active and important to these key species.
Commissioning (budget/procurement)
This project was facilitated through the expansion of an existing relationship with the Institute of Zoology (IoZ). The budget range for this trial was less than £50K and lasted a year.
Deployment (what / who / where / how long)
In Richmond, the cameras were set up in small clusters in selected locations for 3-4 week deployments. There was a minimum of nine deployments of at least 20 cameras undertaken between autumn 2021 to autumn 2022. Three weeks was sufficient to gather information on the species present or migrating at a location, whilst avoiding significant duplication of data.
The Institute of Zoology managed four of these deployments, from camera set-up through to analysis of the data, using their self-developed automated image recognition using machine learning technology. Officers and citizen scientists managed the other deployments. The cameras were placed wherever it was needed to capture images of passing wildlife. This then built up a picture of which species were present in the monitored areas, and gave an indication of relative species abundance.
The Kingston team conducted surveys using state of the art wildlife cameras deployed across various sites in 53 locations, from April to August 2023, capturing animal activity throughout nature reserves in the borough of Kingston. Kingston council staff, working with the London HogWatch Project, as well as volunteers, facilitated camera placement using GPS coordinates generated by a computer and Google MyMaps. All cameras were in place for around 2 weeks per site.
Cameras were set to trigger and take a photo every second if an animal entered the detection zone of the camera. Use of infrared flash meant cameras could also be active at night. To ensure even coverage of each site and to align the Random Encounter Model 7 protocol, cameras were placed as close as possible to a predetermined grid pattern. All species in a 24-hour period which triggered the camera were ‘tagged’. The processed data was then used to calculate trapping rates (amount of sightings/number of nights the camera was active) for each site and species.
Early indications from the cameras in the field (from which photos can be manually extracted) produced an average of around 30 images per day. The cameras were set on a three-shot multiple mode to increase the chances of getting a clear shot of the animal. As most wildlife activity is in the lowest traffic areas, this is where the cameras were deployed as a priority. Cameras were moved monthly depending on data needed for specific projects as well as to cover the expanse of land that data is sought for. At busier times of year for animal activity, as little as two weeks might be enough to determine the presence or absence of a species and a rudimentary analysis of relative abundance.
Technology Implemented
The suppliers of the sensors were provided by Wildlife Future Outdoors Ltd. The trial deployed a number of wildlife cameras and motion sensors connected to a cellular adapter. The motion sensor detected movement and triggered the camera to take images. The cellular adapter then transmitted the image data back to a dashboard in the cloud, tagged with date, time, and location data. The data was then accessed by the Zoological Society of London (ZSL). Images were manually tagged by their team and their volunteers. The manual identification was then fed back to the data processors so the AI could “learn” to identify different species based on this information, ultimately taking over the process of identification entirely.
Although there are all-in-one units that combine the camera and cellular technology in one device, the image resolution is typically of a lower quality. Officers decided to buy a good camera and then a separate universal cellular adapter which is attached to it and holds the memory card. These items come from different manufacturers. Kingston used Browning Strike Force Pro camera traps.
Results / Key Findings
Overall, the results of the data generated were positive. One of the species of greatest interest was a protected species. Sensors were finding more animals of this species than expected. The trials gathered data from 220 camera placements between the spring and early winter. Images of hedgehogs, a bathing kingfisher, grass snake, the rare water rail bird and a muntjac deer with young, were all captured. Data collected from the Kingston surveys enabled the calculation of trapping rates for different species at each site. Additionally, various mammal and bird species were recorded, including those of conservation concern.
Benefits / Usefulness of Data
The Richmond trial sought to gather information on the presence or absence of key species in the area, to inform plans for the management of council-owned land. A good number of various mammal and bird species were recorded, including those of conservation concern.
Kingston concluded that this was the most comprehensive wildlife camera survey that has ever been conducted in the borough, and will inform conservation management of our nature reserves in the future. One of the largest insights was as to the scale of dog disruption to animals and how with considered fencing there could be a dramatic reduction in dogs interfering in areas that the Council wish to encourage biodiversity for all the reasons noted previously.
Lessons Learned
It was clear that officer engagement is imperative. Placement, and quality, of the cameras is crucial. Significant time could also be saved by using AI to support data processing.
Conclusion
The South London Partnership’s IoT trial has proven to be a significant step forward in monitoring and conserving biodiversity within Richmond and Kingston. By leveraging advanced camera and sensor technologies, the project provided real-time data on local wildlife, which was previously challenging to obtain through traditional methods. The deployment of cellular camera traps allowed for extensive monitoring with minimal human interference, capturing valuable insights into species presence, abundance, and behaviour across key conservation sites.
The collaboration between local authorities, the Institute of Zoology, and volunteer citizen scientists was instrumental in the project’s success. The data collected not only highlighted the presence of protected and cryptic species but also shed light on the impact of human activities, such as dog disruptions, on wildlife. This information is crucial for shaping future conservation strategies and land management practices.
Overall, the IoT initiative has demonstrated the potential of technology in enhancing environmental conservation efforts. The positive outcomes, including the identification of key species and the generation of comprehensive wildlife data, underscore the project’s value in promoting biodiversity and resilience against climate change. This case study serves as a model for other regions looking to integrate IoT solutions into their environmental monitoring and conservation programs.
Photos from the use case
Contact
For further information, please contact the service leads involved in this project, listed below.
Elliot Newton
Biodiversity Officer
Royal Borough of Kingston
elliot.newton@kingston.gov.uk
Bethany Pepper
Programme and Policy Lead (Climate Change and Sustainability)
London Borough of Richmond
bethany.pepper@richmondandwandsworth.gov.uk
Rebekah Brown
Business Analyst
Royal Borough of Kingston
rebekah.brown@kingston.gov.uk
Pierre Venter
IoT Delivery Manager
Royal Borough of Kingston and London Borough of Sutton
pierre.venter@sutton.gov.uk
