VICAS
Distributed Fiber-optic Coherent Reflectometer as Vibrational (and Temperature) Sensor
The System
VICAS uses standard single-mode optical fiber (G.652) as highly-effective distributed sensitive element detecting soil vibrations (acoustic vibrations) at a distance of several dozens of kilometres along the optical cable (longest applications has been 70 km, currently tending to 100+ km). VICAS system (hardware and software) detects, localizes, classifies and alerts on activity along the entire length of the fiber-optic cable. It offers cost-effective method of around-the-clock automated environment activity monitoring over long distance linear objects.
VICAS allows to detect various events with 10 meters accuracy throughout monitored zone. Detection accuracy is greatly increased by a proper zone segmentation. Last but not least, VICAS can be interfaced to pre-existing optical cables even if posed inside tubes; with an appropriate tuning deratings in performance are well acceptable while deployment costs drastically reduced.
Since VICAS permits to detect any kind of activity that is causing water or soil vibrations, it can monitor approaches to a guarded zone, movement along the border or trespassing of the monitored zone border. Therefore, If any person, animal or vehicle approaches to guarded zone or any active process starts near protected object, VICAS generates and forwards signal events (i.e. alarms) to the control room indicating the source location.
Although some customizations would be possible, VICAS is allocated in a 3U chassis to be installed in a standard 19" telco rack. The chassis has three slots for: receiving/transmitting unit (coherent reflectometer), amplifier module, and industrial computer for data processing and data communication with remote server. VICAS system can operate independently or may be integrated to any existing security system. Open API allows developers to connect VICAS to any compatible security (and safety) application using Ethernet communication channels.
Physical working principle
VICAS operation is based on physical principals of coherent reflectometry. Coherent reflectometers differ from common reflectometers by using highly stable and coherent laser sources. Optical pulses are repetitively launched into the fiber and during their propagation along the line a small portion of light is scattered backward. This back reflection occurs both from fiber local defects and also from random microscopic fiber index of refraction non-uniformities (sometimes called "scattering centers"), uniformly distributed over the fiber length. The backscattered optical waves are summing up coherently, with fixed yet random phases thus forming a chaotic intensity signal. This signal is well reproduced from pulse to pulse, as long as the fiber and laser remain intact.
Subtle fiber deformations caused by vibro-acoustic events change intensity of the back scattered (or reflected) signal. Most basic analysis of the acquired data consists of maximum difference calculation for subsequent measurements of the backscattered intensity which is performed for every location along the fiber. The analysis of interferometric pattern variations contained in the backscattered signal allows to reveal both locations and characteristic features of vibro-acoustic events affecting the fiber.
Event recognition
A highly essential component of VICAS is its embedded signal recognition algorithm. Signal processing is performed by using a convolutional neural network. Each signal acquired from a specific fiber segment by the optical equipment is then assigned with a "probability vector" which is compared to the etalon reference signals. Having performed analysis of the acquired information over the whole line, the recognition module algorithm aggregates probabilistic events into an object or set of objects. At the final stage, specific object characteristics (trajectory, velocity, etc.) are determined.
Implementation of neural network, capable of learning to new types of events, allows to widely expand a range of classifiable objects and solvable tasks. The neural network software can run either on the built-in computer of VICAS module or on the external server. In case when multiple VICAS modules are operating the software is better deployed on the external server.
The sketch about event recognition sensitivity is purely indicative since largely affected by the mechanical impedance between the cable and the source of vibration. It can be significantly increased by properly placing and burying the optical cable. Vice versa, a derating is expected by using pre-installed cable in case they are piped.
Applications
Because of the VICAS' capability to catch a signal dozen meters far from the sensing cable, “early warning” of critical infrastructures is one of the main applications. Thus, “Border Control” is the most appealing one.
Other typical areas of application of VICAS are: "Pipelines monitoring and diagnostics" (Routine works monitoring and protected zone monitoring); "Transportation infrastructure objects monitoring" (Rail tracks, landing strips, highways, bridges, tunnels); "Communication lines guarding".
Currently, a number of applications based on “modal analysis” are under testing with the aim to monitor structural misbehaviour of large infrastructures (i.e. bridges, dames, tunnels, etc.). For newly built structures, taking advantages from experience in aeronautical structures, where the technology is well engaged, it is also foreseen the possibility for VICAS to monitor “fatigue cumulation”.
