top of page

AquaWatch AEI Revolutionises Water Monitoring with AI Cameras and Active Learning

The AquaWatch Environmental Intelligence (AEI) uses an active learning training system to quickly train custom image classification models to enable complete control of training models, with classes agreed with the customer, and fine-tuning of models as additional data becomes available. This modular approach enables layered solutions alongside the specific CSO incidence monitoring. The capabilities of the AI Camera provide the ability to measure water levels, identify rubbish, measure leaf velocity / speed of flow, detect floating elements, and identify pollutants such as oil on water. Monitoring algorithms can also complement the physical water parameters gathered in real time by the AquaWatch Waka, adding to a more complete picture of the health of waterways in real-time.

Here are some examples of how AI Cameras can be used to monitor and recognise key environmental indicators:

Rubbish Identification

AI cameras can be trained using machine learning algorithms to recognise different types of rubbish such as plastic bottles, bags, or other debris commonly found in water bodies.

Training data for rubbish identification can include a wide range of images showing various types of rubbish in different environmental conditions. The AI model learns to distinguish between natural elements and rubbish by analysing visual features such as shape, color, texture, and context.

Leaf Velocity Measurement

AI cameras equipped with machine learning algorithms can monitor the movement of leaves on the water surface to assess the water's flow dynamics and overall health.

By analysing the motion of leaves using computer vision techniques, the AI system can estimate the velocity and direction of water currents. Machine learning models can also be trained using annotated data that associates the movement of leaves with known water flow characteristics.

Floating Debris Detection

Similar to rubbish identification, AI cameras can employ machine learning to detect and track floating debris such as leaves, branches, logs, or other objects.

Machine learning models are trained to recognise the visual patterns associated with floating debris, distinguishing them from other elements like waves or natural features. Training data for floating debris detection includes images and videos of water surfaces containing various types and sizes of floating objects.

Oil/Pollution Detection

AI cameras can be partnered with machine learning algorithms to detect oil slicks and other forms of pollution on water surfaces.

The machine learning models can be trained to recognise the unique visual signatures of oil slicks, which often exhibit distinct colors, textures, and patterns.

Training data for oil/pollution detection includes images and samples from areas with known pollution incidents, as well as simulated data to cover a wide range of environmental conditions.

In all these cases, continuous monitoring by AI cameras equipped with machine learning capabilities enables real-time assessment of water health parameters, allowing for timely intervention and remediation efforts to mitigate environmental impact and ensure the sustainability of aquatic ecosystems.

Contact us to find out how the AquaWatch AI camera and systems can provide valuable data and insights for research and environmental management initiatives.



bottom of page