We present the results of laboratory experiments on a down-scaled river levee constructed with clayey material collected from a river embankment where a permanent resistivity instrument has operated since 2015. To create potential seepages through the levee, two zones (5 × 4 cm and 10 × 2 cm) were filled with sand during the levee construction. Electrical resistivity tomography (ERT) technique and Fiber Bragg Grating (FBG) technology were used to study time-lapse variations due to seepage. The ERT profile was spread on the levee crest and the Wenner array with unit electrode spacing a = 3 cm was used. Six organic modified ceramics (ORMOCER) coated 250 μm-diameter fibers were deployed in different parts of the levee. Time-lapse measurements were performed for both techniques from the beginning of each experiment when water was added to the river side until the water was continuously exiting from the seepage zones. The results showed that ERT images could detect seepages from the early stages. Although with a short delay compared to ERT, fiber optic sensors also showed their ability to detect water infiltrations by measuring temperature changes. Both technologies being successful, a discussion about respective peculiarities and pros and cons is proposed to suggest some criteria in choosing the proper technique according to the specific needs.

Optical fiber sensors based on an interferometric approach appear attractive for an extensive application of optical sensing in the pervasive fiber infrastructure, already deployed for telecommunications purposes. In this paper we show the performance of a sensing system based on Michelson interferometer, exploiting a 48-fiber telecom cable in a conduit under the sidewalk running alongside 4-km railroad tracks, installed by an Italian provider in the north side of Lombardia in Italy. The proposed interferometric scheme does not require an isolated reference, taking advantage of the geometrical arrangement of the sensing and reference fibers inside the same cable to cancel the strong common mode noise, accumulated in the railway hostile environment. Thanks to the installed sensing system, we monitor the traffic, identifying in a very simple way the passage of trains and the presence of cars at the railroad crossings along the railway. Moreover, we trigger events potentially dangerous for the railway, such as the fall of heavy rocks from the walls along the rail tracks. Thanks to an appropriate combination of features, we achieve an effective and robust real-time event classification by a supervised artificial neural network, able to recognize the rockfall as a dangerous event for the railway integrity and provide a prompt alarm, minimizing the nuisance false alarms due to the environment. The proposed sensing system embedded in the telecom network provides a sustainable solution, in terms of cost, energy efficiency and reliability, without the necessity of coherent detection, high speed sampling, and complex digital signal processing.

The results of an experimental investigation of the impact of the bending-induced linear birefringence on a fiber-optic current sensor are presented. A polarimetric configuration is used for measuring a 50-Hz AC current, exploiting the Faraday effect in a coil of standard optical fiber. Different values of diameter and number of turns of the coil are considered, evaluating the loss of sensitivity due to the birefringence as reduction of an effective value of the Verdet constant. It will be experimentally demonstrated that for a coil with a diameter sufficiently large, i.e. equal or larger than 14 cm, the effect of the linear birefringence plays only a marginal role and it can be neglected.

In this work two Neural Network (NN) based solutions are proposed to recover the distributed temperature profile of a sensing fiber, measured using a commercial Brillouin Optical Time-Domain Analysis (BOTDA) interrogator. A detailed analysis in terms of temperature accuracy and processing speed is carried out for both the proposed methods, comparing the results with the ones obtained from the application of classical fitting techniques, namely cross-correlation (CORR), Lorentzian fitting (LF) and Pseudo-Voigt fitting (PV), through both simulations and real measurements carried out in laboratory environment. The results show that the first NN implementation, which aims to maximize the accuracy of the temperature profile and the processing speed, can handle different width of frequency acquisition window but needs to be optimized for a specific frequency acquisition scanning step. The second NN implementation, however, can also handle different values of the acquisition scanning step with a minor performance drop. Simulations and experimental data show a massive advantage of NN implementations in terms of processing speed with respect to classical fitting techniques, with a slightly better accuracy of the estimated temperature profiles.

In this letter, a novel implementation of a polarimetric fiber-optic current sensor scheme based on polarizing beam splitter and a $22.5^{\circ }$ Faraday rotator will be presented. The developed sensor employs a compact and pigtailed component realized by joining a polarizing beam splitter and a Faraday rotator, it does not require any external adjustment to work and shows very high immunity to external electromagnetic and mechanical interference. The experimental results prove good sensitivity of the sensor allowing for the measurement of small value of current down to 100 mA.