Description of the project
Landslides are one of the most frequent geological disasters. They represent a serious threat to human lives, as well as a huge cost in terms of infrastructure damage. Most of them are shallow, i.e., the sliding surface is typically located to a depth from a few decimetres to some metres. In the past, the approach to addressing the issue of shallow landslides was more defensive than preventive.
SlopeGuard is a novel approach that aims to change the way such landslides are monitored, managed and prevented. Our approach is unique and effective since it integrates a bespoke monitoring device (SlopeGuard-ROD) with advanced machine learning techniques (SlopeGuard-Brain). This mutually reinforcing combination of tailor made HW and SW allows SlopeGuard to generate early warning of landslides 24/7 in a fully automated way.
Mid-term project update
As part of the SlopeGuard project, aimed at predicting shallow landslides, the consortium completed the conception, design and construction of the embedded monitoring station “Slopeguard” that will be used to for the supervision of active shallow landslides. This is a flexible but durable sensorized rod designed to be easily inserted directly into the ground at a depth of approximately 2m. The data collected by its embedded sensors (accelerometers and soil moisture) will be automatically transmitted to a dedicated web platform for real time monitoring, analysis and storage.
At the same time, the development of machine learning algorithms to predict shallow landslides has begun. Since the data collected by the SlopeGuard sensor are still partial, the consortium teamed up with geology experts from the Italian Alps to gain access to a huge dataset of landslide events that will be used for the development and refinement of the algorithms.
Final project update
The development of machine learning algorithms for anomalies detection and – in the future – for shallow landslides early-warning, has been completed and integrated with the data acquisition system. The AI algorithm has been trained and refined thanks to the data directly collected
by the SlopeGuards activated during the project campaign, together with Techcom landslide dataset and other geoscience datasets provided by supportive customers.
The Slopeguard system brings the flexibility of an embedded monitoring station and the innovation of integrated AI into the market of landslides monitoring solutions.