Innovative Robotics Project Aims to Revolutionize Environmental Monitoring

Strategic Insights -

Botanical surveys are crucial for detecting and tracking the impact of climate change across different ecosystems, but current methods are often slow, tedious, and costly. A recent robotics project by Falmouth University student Cas Penfold has proposed an innovative solution: a dog-mounted computer vision system that can make environmental monitoring both cheaper and faster. The project explores how monitoring plant species is vital for identifying harmful botanic and detecting climate change.

Key Takeaways:

  • Cas Penfold's dissertation project, a third-year Robotics BSc student, examines the limitations and importance of capturing botanical surveys using a dog-mounted computer vision system.
  • The project aims to create an alternative way to monitor plant species, increasing the amount of surveys that can be taken place with lower costs.
  • Botanical surveys are crucial for detecting and tracking the impact of climate change across different ecosystems.
  • Current methods of botanical surveys, such as using detection dogs, have a high initial cost for specialized training ($15,000-30,000 Australian dollars each).
  • The project successfully trained a computer vision model to identify sunflowers in frames of collected videos with a high level of accuracy.
  • A bespoke harness with a gimbal and small action camera was created to capture footage from the dog while it explores the environment.
  • The project demonstrates the potential of using dogs as a resource for environmental monitoring, making it more accessible and cost-effective.

Statistics:

  • The initial cost for training detection dogs can range from $15,000-30,000 Australian dollars each.
  • The project aims to make environmental monitoring more accessible and cost-effective by utilizing dogs that are walked off the lead in rural environments.
  • The computer vision model was trained on thousands of hand-labelled supervised training images of sunflowers.
  • The project achieved a high level of accuracy in identifying sunflowers in collected videos.

Sources:

  • Cas Penfold's dissertation
  • Falmouth University