Student intern: Olivia Ortlieb (Computer Science, Philosophy)
Supervisors: Loraine Clarke (School of Computer Science), Simon Dobson (School of Computer Science), Alison Johnston (School of Mathematics and Statistics), Martin Skelly (Duncan of Jordanstone College of Art and Design), Harry Watkins (St Andrews Botanic Gardens)
The project set out to explore interesting and playful ways biodiversity that is usually unnoticed can be captured using sensors and technology with a larger agenda of engaging the public with biodiversity through tangible interactive experiences that highlight biodiversity activities and changes. Specifically, how can sensors and technologies capture unnoticeable biodiversity changes living on and in deadwood at St Andrews Botanic Gardens? We wanted to explore how we could use low-cost DIY technologies such as Arduino or Raspberry Pi to start observing the changes in biodiversity. The end goal is to later create interactive digital experiences that encourage the public to notice biodiversity that is typically unnoticeable. First, we needed to understand what unnoticeable changes in biodiversity were happening.
The most exciting finding was observing the pile of sawdust that increased on a daily basis as a creature appeared to be burying into the wood above the pile.
Camera Prototype:
The camera prototype consisted of a raspberry pi 3 model V1.2, High Quality Camera Module – M12 Mount, a 15MP 2.7mm focal length M12 wide-angle lens and a power bank (16000mAh, 59,2 Wh) packaged up into a lunch box with a hole in the side of it for the camera lens to be put through. The prototype was placed on a laser cut wooden tripod stuck into the ground to keep it in the same place every day (Figure 1). The camera took photos of the same area of the dead wood hedge to observe the changes over a 4-6 week period in the evenings and overnight during the summer months.
Data Captured:
It was really exciting to see a range of changes in biodiversity captured from the camera such as, frogs climbing through the wood pile and moving logs, a pile of sawdust beginning to form and increase in size on a daily basis indicating a creature burrowing into the wood logs (figure 3), slugs, flies, ants, long thing small insects and very small insects we couldn’t identify (figure 2).
Microphone Prototype:
The microphone prototype consisted of a raspberry pi 3 model V1.2, the waveshare WM8960 Audio HAT and a power bank (20000mAh, 74Wh) enclosed in a lunchbox with the microphone outside of the lunch box in a zip lock bag to be semi-waterproof. The prototype was placed at the wood pile with the microphone placed in the wood pile (figure 4) taking audio recordings in the evenings and overnight during the summer months.
Data Captured:
The prototype picked up the sounds of several different creatures and activities at the wood pile such as scratching, tapping, buzzing, croaking and what sounds like the chirping of crickets (figure 5).
Some reflections from Olivia, the BIRCH student intern
Hello! My name is Olivia Ortlieb and for my Birch internship, I investigated using sensors to monitor ordinarily-hidden biodiversity activity in deadwood in the Botanic Gardens. I primarily did this through the use of camera and audio systems. After a lot of experimentation with different cameras, I ended up capturing images using an M12 camera mount connected to a Raspberry Pi and placed in a lunch box on a laser cut stand. I also used a soundcard placed in the deadwood and connected to a Pi via an extension cord to record audio.
Each day, I inspected the captured content and logged indicators of biodiversity, including the sounds of buzzing and crickets, photos of insects, and the growth of a pile of sawdust likely related to wasp activity. I elevated my data logging by using a script to highlight likely insect activity by comparing each photo with a set of ‘default’ images of the same location. I also leveraged audio post-processing to make feature identification easier by reducing noise and narrowing in on specific frequency ranges.
Towards the end of data collection, I compared the sightings of different species to other variables like the temperature, humidity, rainfall, and time of day to see if any patterns were present. I found a few correlations like a sharp decrease in incidents of buzzing with the approach of sunset as well as during rainfall. I also captured the deadwood site using 3D photogrammetry techniques.
This project was a fantastic opportunity to gain deeper insights into biodiversity monitoring while honing my technical skills. I’m excited to see how these methods could be refined and expanded to further unveil often overlooked ecosystems.
Interactive Experiences with Biodiversity presentation by Olivia Ortlieb