EPCAPE-PIERSEDGE

A software-defined sensor makes use of algorithms inferring information near the source of the signal. This is known as edge computing. An example that will be explored in this field campaign is the use of steerable thermal cameras. Early experiments have shown these camera systems (repurposed security systems) can delineate between different cloud layers. We will explore different ways of scanning the camera system in azimuth and elevation to map out cloud fields and stitch images to create a “Total Sky Imager-like” product. We will also look at how scanning should adapt to cloud fields. For example, in clear conditions perhaps the system should dwell on the sea surface instead. This field campaign will deploy a Sage Waggle[1] edge computer device and a Mobotix steerable thermal camera to the Scripps pier. Use of AI at the edge will also be explored for applications like biodiversity through avian and other species detection and the use of computer vision to detect possible anthropogenic aerosol sources such as boats and vehicles. All code is open source[2] and data will be available in near-real time.[3]

[1] https://docs.waggle-edge.ai/docs/about/overview 

[1] https://github.com/sagecontinuum

[2] https://portal.sagecontinuum.org/nodes  

Timeline