1. Magnetic Ordinance Detection
By Christopher Fenton

2. Goals
Analyze feasibility of magnetic ordinance detection methods, specifically with IED detection in Iraq in mind
If feasible, build working prototype
Successfully detect something metallic

3. Different Approaches to Object Detection
Traditional Metal Detectors
Ground-Penetrating Radar
Magnetic Detectors

4. Magnetic Detection Approaches
Balanced-Loop
Detects change in B-field over time
Covers large areas
Magnetometers
Measures absolute B-field
Covers small areas

5. Balanced Loops
First use of Magnetic “Indicator Loops” for harbor defense in 1915 by British in WWI, adopted by U.S. in 1942 during WWII
Can only detect moving magnetic disturbances
Typically large and immobile (>1.6 km^2)
Abandoned for harbor defense in favor of SONAR following WWII

6. Balanced Loops in Action
Old detector station in Nahant, MA

7. Magnetometers First invented in 1833 by Carl Gauss
Can detect magnitude and direction of magnetic field
Small and lightweight
Still used for geological surveying and “Magnetic Anomaly Detectors”

8. Magnetometers in Action
Magnetometer Array used for UXO detection
Circuit model of sensor used in MicroMag3 (Sensor inductance changes with external B-field)
MicroMag3 3-axis Magnetometer

9. Approach: Magnetometer Array
Sensors are small (~1”x1”), cheap ($50) and easy to handle
> Even small loops are several m^2
Insensitive to scanning speed and tilt
> For loops, tilt and speed need to be precisely monitored
Arrays can be scaled to arbitrary width for wide-area scanning
> Magnetometers give point measurements, but can be expanded to cover wide areas like loops do

10. The MAGNETube

11. MAGNETube Setup 3 x MicroMag3 3-axis SPI magnetometers
Sensors mounted 15” apart
Calibrated so Earth’s B-field = 1 = G
2 x Picaxe 18X microcontrollers
Expandable through “daisy-chaining”
1 Laptop running “Listener” software and outputting to CSV file for analysis in Microsoft Excel®

12. Setup A B C

13. How is the magnitude computed?
1. X, Y, and Z values for all 3 sensors are sent to laptop
2. Calibration offset is subtracted from each direction
3. Magnitude = √(X^2 + Y^2 + Z^2)
4. Magnitude is scaled from range to approximately equal “1” in Earth’s B-field
5. Sensor: 1= Gauss in Los Angeles

14. Test 1: 80 lbs of Iron Location: Erdem’s Apartment
Target: 80 lbs of iron weights in a plastic trashcan

15. Test 1: 80lbs of Iron
Possibly due to misalignment of sensor during test
Conclusion: Readily detectable if directly above pile, drops off quickly

16. Test 2: 4” Brass Artillery Shell

17. Test 2: 4” Brass Artillery Shell
Background: 12” above ground
Test 2: 12” above ground
Conclusion: Brass has no magnetic signature. Only bolts were detectable, and only then at close range.

18. Test 3: Neodymium Magnets (high sensitivity simulation)
Large 3”x6” Neodymium magnet

19. Test 3: N.D. Magnet
Conclusion: Magnet is easily detectable at a reasonable range

20. Test 4: Attenuation in Water

21. Test 4: Submerged N.D. Magnet
Conclusion: Water has no attenuation effect on magnetic field

22. Future Improvements
Use faster microcontroller with on-board FPU (~3X improvement in sampling rate)
Add wireless serial link for easier calibration and field-use
Experiment with distortion detection vs. simple magnitude detection
Use higher-sensitivity magnetometers and higher-density array
Compare vs. traditional metal detector

23. Conclusion
Undocumented hardware failure-modes can be extremely difficult to fix
Magnetic detection appears to be a valid method (and is apparently in-use)
A simple array can be constructed for less than $250
With more time, the current design could be greatly improved