Introduction

Technology has rendered power a concept much like the air we breathe today; it often goes unnoticed unless there is not any. Life has become so dependent on it that every flick of a switch brings the expectation of illumination - and no compromises. We even go as far to power our alarm clocks from the wall outlet, with simple trust on uninterrupted power so wakeup services will be performed right on time, every day. (Companies have policies that employees are excused of being late if there has been a major blackout, like the one in New York recently). It began a century ago by extending a few wires to get the porch light going, nonetheless, power is no longer considered a mere flow of electricity. Today, power is the flow of information, the flow of life.

US - A Power Substation, whose damage would black out an entire city

Please see some examples below, about what can happen to power lines that are veins of civilization.

Damage 1, Damage 2, Damage 3, Damage 4, Damage 5

Not only dependability of power is undeniable, it is becoming increasingly essential every day. And the veins of power distribution, is composed of the power grid. Protecting the 157,810 miles of power lines in the US (many in very remote areas and with no way to monitor them) against possible terrorism, and natural disasters is the main focus of this project. Once established, power lines can be monitored automatically 24 hours a day, and problems can be spotted immediately - sometimes even before they happen. (Power companies can tell if something has happened to the lines, but they will need to find where it happened, and actually go there to know what happened). Our goal is to design this system affordable enough, so that developing countries could also benefit from it.

Conventional control and protection mechanisms on power transmission networks are blindsided in recognizing and responding to physical threats to the power system including but not limited to sabotage, and acts of nature. These safety mechanisms are based purely on electrical quantities and the operator will not have knowledge of attacks on the system that physically alter the transmission path, which is considerably common occurrence. A distributed computerized automatic early warning and control system via wireless sensor technology to assess the mechanical health of power transmission hardware is a promising design for achieving a higher level of survivability for power transmission system as a whole. In addition to immediate threats, potential threats as well could be detected allowing counter-measures to be taken in the earliest manner. A sensor network that monitors the energy network in such a way that the sensors can make a wide array of measurements, which can be analyzed, and proper preventive and/or remedial actions can be taken at appropriate times. The architecture will be self diagnosing so that a small number of traitors (i.e. faulty equipment) may not lead to a false alarm and/or incorrect/inappropriate/unnecessary decisions. A broad number of environmental and human related events can be detected within a matter of seconds and countermeasures can be taken/suggested. Relations between the measurements can be exploited for further fault classification. Specialized sensors called camera nodes as a part of this proposed network add it the ability of interpreting video and make decisions based on whether a hostile situation (i.e. sabotage) is happening in that particular scene. In such case that particular frame is captured and logged, an alarm is then raised to be propagated via the wireless sensor network, containing information about the location and possible type of activity.

Feature Extraction is a major area of interest in computer interpreted vision, with a variety of uses ranging from satellite surveillance to medical imagery. The main idea is if we could represent images in mathematical terms, we should as well, be able to represent the features in them in the same manner (i.e. lines, edges, curves, geometrical shapes, motion, color changes, light intensity, shadows) and thus be able to extract them from the image as necessary. We can also tell the approximate size of these objects, their speed and direction. In this project we are also developing a specialized network using hardware-software co-design techniques, whose dedicated purpose is interpreting a digital motion picture and making decisions based on whether a certain criteria is being met in that particular scene. This criteria about what constitutes a normal (i.e. acceptable) scene, is known a priori and corresponding parameters are supplied to the computer. (There are cases in which the computer can learn the good from bad by itself). Based on these the computer interprets image sequences and attempt to come to a logical conclusion whether there is a hostile situation in the picture. If so, that particular frame is captured and logged. An alarm is then raised to be propagated via the wireless sensor network, containing information about the location and possible type of activity. There are several uses for such system, such as on-the-move traffic sign detection, public surveillance (i.e. detection of a weapon by its shape) et cetera. This study is based on monitoring structural integrity of objects - power transmission lines and hardware in particular. “Intelligent” Cameras placed in vicinity of transmission lines constantly monitor their crucial parts, whose dedicated computer interprets their image and takes necessary action(s) in events like a wire snap, a pole collapse, or a live being walking into this dangerous area – ranging from raising an alarm to cutting the power. Transmission lines being inherently dangerous, make it beneficial monitoring them from a safe distance, ensuring their continuous operation, and rapid detection of problems without an operator on duty.

Also, On The News:

 http://www.newscientisttech.com/article/dn9482-wireless-sensors-could-protect-us-power-grid.html

http://www.iastate.edu/~nscentral/news/2006/jun/grid.shtml