Smart Edge-AI Camera

Cameras mounted on poles, farm structures or vessels capture video and still images of cages, gear handling, vessel movements and surface conditions around aquaculture sites. These views record activities such as stocking, grading, cleaning, harvesting, gear maintenance and boat operations that conventional water quality sensors cannot describe. Edge artificial intelligence models running close to the camera process this imagery in near real time, converting it into structured events and metrics rather than just storing raw footage.

Machine vision models identify and classify above water events and features, such as when gear is lifted or cleaned, when vessels arrive or depart, when equipment is moved, when fouling or debris on structures increases, and when product moves across grading or processing conveyor lines. They can be trained to detect and track individual objects, estimate size and count, and recognise specific classes of product or equipment. Edge artificial intelligence then summarises these detections on the device, sending compact event records, counts and timestamps back to the cloud instead of continuous high bandwidth video. Each event can be geolocated and time stamped, then fused with environmental and operational data to build a detailed history of actions and conditions at the farm.

In the data dissemination layer, events derived from above water cameras are displayed alongside sensor time series and other operational records on dashboards and time lines. Users can review how site conditions and outcomes relate to activity patterns, seeing for example when cleaning runs occurred relative to changes in fouling indicators or performance. Rules and thresholds can trigger alerts when the camera system detects missed or abnormal activity, such as extended periods with no maintenance passes, unexpected vessel presence or unusual handling, so managers can respond promptly.

By feeding machine vision and edge artificial intelligence outputs from above water cameras into the same tools as other monitoring data, farm teams gain a clearer view of how day to day practices at the surface influence overall performance and risk. They can refine operating procedures, demonstrate that required tasks have been carried out and maintain an auditable visual record of key events, supporting traceability, reporting and continuous operational improvement.