Magnetometers are one of the basic instruments used for the detection and discrimination of unexploded ordnance (UXO). Cesium (Cs) vapor atomic magnetometers are commonly used because their readings are independent of the orientation of the sensor, which eliminates the noise problems due to rotation or even vibration of other types of sensors. In order to better discriminate UXO from clutter or scrap, a high spatial density of readings is desirable. Arrays of sensing elements would be useful to efficiently make such measurements and to position such measurements accurately. Existing Cs vapor sensors, however, are large and have high power consumption.

The objective of this project was to develop laser diode technology and use Micro Electro-Mechanical Systems (MEMS) techniques to miniaturize the components and reduce the power consumption of the Cs magnetometer.

Technical Approach

Researchers investigated a high sensitivity, total field magnetometer that was extremely low power, small in size, and capable of being mass-produced for low cost. This research built on previous Defense Advanced Research Projects Agency (DARPA)-funded efforts to develop chip-scale atomic clocks (CSAC). Cs total-field magnetometer technology is similar to the clock technology used in CSAC.


The project team initially tested Vertical-Cavity Surface-Emitting Lasers (VCSELs), identified a promising cell filling methodology, and identified several interrogation methods. The team then focused on real-world scenarios, building physics packages, laser diodes, and measuring and optimizing the performance to determine a final design. The team continued to explore the complicated design space and understand the effect on cost and performance of the various components in order to implement an optimal design. The team built two generations of prototypes, which prove the capabilities of miniature sensors. Finally, the team re-designed the electronics for a new generation of improvements in size and performance.

Significant accomplishments included constructing prototypes with sensor power less than 100 mW (excluding readout electronics), building sensor physics package size less than 1.0 cc (excluding electronics), building electronics occupying approximately 15 in2, demonstrating potential production cost of about $200 per sensor in large quantities, and demonstrating sensor sensitivity of 5 - 15 pT / root-Hz.


Smaller sensors could greatly improve the field performance capabilities of total-field magnetometers. Whether single sensor configurations, gradiometers (dual sensor), or arrays of larger numbers of sensors are desired, products built from these new sensors would be less expensive and easier to use, translating into faster, and thus cheaper, UXO surveys. Discrimination capability also will be enhanced by a higher density of measurements with greater positioning accuracy. Small real-time output devices may be built using these devices, either in wands or in lightweight arrays. Such systems allow for data analysis and output in the field, providing real-time information on the parameters of the target. Arrays of sensors may be deployed on small platforms such as unmanned aerial vehicles.