Perim-Alert®

[click on the image above to view larger]

I was able to score a case of Perim-Alert SR-1 Sensing Units from one of my obtainium sources and have been looking for a cool (and useful) way to incorporate them in a perimeter security system.

The units are brand new ( I’m guessing they were built in the mid 80s?), in-the-box, and come complete with hose clamps that are used to mount the units on standard galvanized fence posts.  These units, according to the box, were manufactured by Air Space Devices (16624 Edwards Road P.O. Box 7500, Cerritos, CA  90701) and were distributed by Norton Safety Products.  From what I can tell, Air Space Devices is no longer doing too much anymore or  is deeply “off the grid.”

The cool thing about these things is their extreme sensitivity.  The slightest jiggle causes a switch to open and, in conjunction with a latching relay, can be used to monitor and protect any barrier that needs to be accessed (a gate, door, stairway, ladder, etc.) by someone approaching a structure or property line.

When you open the sensor, you see a pretty simple setup:

The orange blob in the center is a vibration sensor that is sealed in an epoxy case.  The normally open (NO) contact terminals on the left and right of the unit are wired into your monitoring circuit.  The vertical screw in the center is used to adjust the sensitivity of the device.  It acts as a simple counterweight.  Loosening the screw makes the lever arm of the accelerometer longer, which, creates a larger moment of inertia to act upon.  Tightening the screw creates a shorter moment, enabling smaller movements of the device to shake the spring steel that the accelerometer is attached to.

The unit is unpowered meaning it requires no external power source to operate.  It is merely a switch.  Movement shakes the unit, and the circuit opens, causing the interruption of electrical flow which can be easily turned into a signal that activates alarms.

The usual installation of these devices would have probably entailed connecting a number of these units around an entire fence line or some type of perimeter structure in a continuous series circuit along with a central power supply.  Thus, a 24V or whatever control voltage would be sent around a loop, passing through each sensor on the perimeter and creating one, large circuit that would be monitored for continuity at a base (or guardhouse)  somewhere.

If any part of the perimeter circuit was either moved or cut, the circuit would activate some alarm and a crew of henchmen carrying machine guns and riding on snowmobiles would sprint out along the perimeter, looking for the intruder.  Evidently, with this type of setup, it would be hard to pinpoint the exact breach in the perimeter unless you used a sophisticated time domain reflection algorithm similar to old-school telephony fault analysis to pinpoint where the break in the circuit occurred.

Fast forward to today, where we can now use low-powered microprocessors and Wi-Fi networks….

My project will incorporate these units to monitor a perimeter as they were designed.  However, instead of hard-wiring each device together in a chain, I want to install each unit independently of each other and have them report back to me if they are disturbed in any way.

Using simple embedded control and an 802.15.4 star topology encrypted wireless mesh, I can monitor each unit not only for movement and/or tampering, but also track which unit is under attack.  The logic is simple.  The hardware is cheap.  The low-power, 3.3 Vdc operating voltage is minimal.   And you don’t have to run hundreds of feet of copper around a perimeter to control the system.  A simple battery powers each unit and radio.   A low-battery sensor alerts me when a unit’s power supply is getting below a safe threshold and needs to be changed.   Even the communication protocol can be set up to “wake on event” and go into sleep mode until it needs to broadcast the disturbance.

I am thinking:

The Perim-Alert Sensor
A PIC18LF4620 MCU (This is an 8-bit processor;  one bit more and you have reached overkill status.)
Some MRF24J40 transceiver/transceiver modules
A MiWi P2P Wireless Protocol
A battery
A couple of other things.

The PAN coordinator can report to either a web-based app that can notify and trigger some security devices (you know, like a loud alarm or a 500MJ 1.5MA homopolar generator or something…) or a stand alone system that can be accessed through a touchpad like a traditional alarm system that sends SMS messages to a mobile device along with a call to a 911 system.  I don’t know yet.

The first hole you could punch in this design is the fact that all networks are vulnerable to attack and wireless networks are the most vulnerable.  Because this system relies on node autonomy as well as self-organization, RF jamming and packet-level intrusion could easily make the system worthless.  I hear ya…

As this design progresses, I will discuss some ideas and real-world strategies to defend against this type of take-down and additionally, talk about how this is applicable to defending our very own Smart Grid from similar attacks.

Details coming…

In the meantime, check out an application note on the Microchip MiWi P2P wireless protocol here.