How it Happened
In late 2005, I was recalled into military service as the Maintenance Approval and Design Acceptance Authority for Land Command, a formation that comprehended 85% of the Australian Army's assets; including those involved in operations around the world.
As a part of my appointment it was necessary I become qualified as a lead auditor through the auspices of SAI Global.  After qualifying it was necessary I perform technical audits on a large number of units as my appointment had been unoccupied for some period of time.
During the course of these audits, I became aware of how little support military units had to perform the mandatory checks required by regulations.  Many of these activities involved such things as spot checks and stock takes to ensure equipment was available and ready for use.  Data collection was a problem in that clerical manpower had been stripped from units in anticipation the centralised stores system, SDSS would significantly improve the productivity of units.  This it did not.

The evolution of this idea was driven by a number of "imperatives".  These were that the symbology have the following characteristics:

• Be novel such that it could be patented. This would allow the enforcement of a licence condition that the NU-ERA bar code may only be used when representing numbers drawn from the dPId unique number registry.
• Represent a 128 bit binary number of which 96 bits would be the "payload".  96 bits of data provides a number range from 0 to 79,228,162,514,264,337,593,543,950,336.  In other words, it would be possible to issue from the dPId registry one million, billion unique numbers per annum for approximately 79,228,162,514,264 years (~79 thousand billion years).  Provided these numbers were efficiently allocated, this range provides for an inexhaustible supply of unique numbers.
• Be highly resistant to dirt and damage.  This was achieved through large-scale redundancy.  The redundant pattern also facilitated the identification of a bar code in the field of view of an image sensor by the image processing circuitry.
  
• Provide a barcode with a low probability of erroneous read.  This would be achieved through using a 32 bit error code.
• Consist only of dots. Being so it could be printed using inexpensive equipment.  Having only dots is particularly suited to engraving items with the bar code using a simplified laser etching system.
  

Figure 1. Example of a NU-ERA 2D Bar Code

Figure 2. Damaged yet readable NU-ERA Bar Code

Figure 2 provides an example of the robustness of the NU-ERA Bar Code.  This robustness is achieved through the use of largescale redundancy, error detection (to determine if one of the arrays actually yields a valid value) and the processor "overlaying" arrays to determine the strongest summated signal from any "dot" making up each of the overlapped arrays.  Other systems such as Reed-Solomon encoding could be used for this purpose and may indeed be a more efficient use of dots but the redundancy also serves a another function and that is to assist the image processor speedily find a bar code pattern in the field of view of the image sensor.

For more detailed information on the NU-ERA technology suite, please click here to download the dPId information paper.