Castellfollit de la Roca village (Catalonia, NE of Spain) stands on top of a basaltic cliff affected by rockfalls that are
causing the retreat of the slope and the risk of undermining the buildings. So far, remote sensing techniques have been applied. They have detected centimetric precursory movements of toppling in basaltic columns, and unnoticed rockfalls of tiny volume as well. However, rockfalls of remarkable magnitude caused by other mechanisms have occurred at the foundation level of some houses, without any precursor movement detected by remote sensing techniques.

In order to gather information from these events, a priority sector has been instrumented with a geotechnical network of crackmeters, tiltmeters and thermistors to monitor the thermo-mechanical behavior of the rock mass. Readings from the sensors have been automated through a wireless network based on Worldsensing nodes and gateway connected by
LoRa technology.

Tiltmeters are all-in-one devices of sensor and logger operating in dynamic mode. This system enables the detection
of events and sending alert messages in real time. To receive reliable alerts for risk management, suitable thresholds
have to be defined, based on the knowledge of the studied phenomena at the particular scenario. This paper presents the system set up and the threshold analysis through the observation of the cliff behavior on the daily and seasonal cycles registered during the first year of operation.