How dPId began

The genesis of dPId is very much a personal tale, so I have written this in the first person. 

I was recalled into military service in late 2005 because of a lack of Chartered Professional Engineers in the Army.  As a part of my duties, I had to carry out technical audits of military units within a formation called Land Command; comprising around 85% of the Australian Army's assets.  It soon became apparent to me that there were considerable savings to be made if it were possible to uniquely identify all items being held on inventory.  At that time, and it is still the same situation today, the Australian Department of Defence only accounts for comparatively few of its assets individually because, to do otherwise, would cost far more than the savings it would accrue through improved asset management.

There is no doubt that if it were possible to track all items individually, it would result in substantial gains in productivity in business and improvement in military capability in the context of Defence.  Consider for example, how simple it would be to perform a stocktake if all articles on inventory could be identified individually.  Consider how it would not be possible for one office to take the chairs of another in order to make up its numbers.  Recounts would be a thing of the past.  Individual identification allows tracking of the life history of any item and would be invaluable in supermarkets for the control and monitoring of all articles with a shelf-life.  It would also be useful in preventing customers defrauding stores with false returns of defective goods.  Imagine, for example, how easy it would be to account for a week's worth of groceries when taking receipt of them after ordering them online.  In the past, the cost and practical difficulties of doing this have been prohibitive.  In the future, there is NU-ERA.

Exploring RFID as a possible solution

Initially, I considered using a very cheap RFID transponder so I contacted friends and fellow engineers, Ivan Curtis and Mick Evans, with whom I had worked in the past and sought their advice. The concept was to build an RFID tag that could hold 128bits of data and cost as little as humanly possible.  The figure of 128 bits was arrived at because it was felt this provided a sufficient range of numbers such that the supply would be, for all practical purposes, inexhaustible.  128bits also happened to be the largest primary key that could be used in a PostgreSQL database at that time.  This made the number ideally suited for database applications where speed of retrieval was important.

After about a month of analysis, my friends and I concluded that the cheapest RFID tag, that is a chip, with antenna and covering on a label of some description, would cost around US 7 cents.  This we concluded made it too expensive for what was required; namely to uniquely identify everything in any organisation.  We then turned our minds to some form of high density non-volatile memory device like, for example, a small segment of a CD ROM.  This then, eventually, led us to considering some form of two dimensional barcode.

The Idea of a Registry from which Unique Numbers could be Issued

From the outset, it was appreciated that, for this idea to be an unqualified success, it was essential there only be one registry of numbers or, where there was more than one, that registries be able to be identified as being a genuine issuer of numbers.  If that were not the case then there would always be a likelihood two identical numbers appearing in a workplace being attached to two different articles.  Ivan Curtis suggested the way to achieve this was through the use of encryption.  The encryption system could be asymmetric or symmetric.  In the case of asymmetric, it would be necessary to use 256bits of data.  With an asymmetric system, clients would be given a public key to decrypt the number in order to verify it was indeed genuine.  With a symmetric system, the key would have to be kept secret but the number could safely be 128bits using AES256.  The method of encryption and the key used would actually give to the number registry its identity.  In other words, it would be possible to have a number of registries; each using a different key and/or algorithm.  The chances of any two registries issuing the same number would therefore be very small but would still be there.  It was therefore decided there could be different registries for different sized numbers, for example, one registry for 128bit numbers and another for 256bit numbers but never would there be two registries with the same sized numbers.

With the advent of the registry, the commercial model of the dPId became clear.  It would sell the service of providing numbers and maintaining a registry to its clients and would do everything possible to assist its customers to use those numbers to their greatest effect.  In performing that function, there would be opportunities for dPId to sell consultancy services.

Unique Barcode Symbology

Early in the project it was decided to create a new type of 2D barcode.  The reasons for doing this were:

• By having a unique barcode, it would be possible to require that it was a condition of the licence that the barcode only be used to represent numbers from the dPId registry.
• The barcode would only be comprised of dots.  In this way it would be easily printable by low cost printers.  It would also be ideally suited for laser etching.  Such a laser etcher could be very simple in its construction compared to one that had to print squares and other special characters used by readers to identify the barcode in the field of view.
• The barcode could have a high level of redundancy and so would have a greater resistance to dirt and damage than conventional 2D barcodes such as DataMatrix, QR Code and PDF417.
• The redundant nature of the barcode coupled with the manner in which values were represented could aid a reader in quickly identifying a NU-ERA barcode in its field of view.  It would therefore be possible to construct a reader that could work at 200 frames per second and read multiple barcodes in its field of view simultaneously.
• As a fall-back, the NU-ERA number from the registry could be represented by common 2D barcodes such as DataMatrix, QR Code and PDF417 as well as by any commonly available RFID tag capable of holding 128bits of data.

Defence as a Customer Prospect

It was apparent to me, as a serving senior officer in the Army, there was no chance of developing this idea within the Australian Defence Force(ADF).  I had already had a long experience of frustration with trying to introduce new ideas into the ADF in the past.  (See Rifle Project)  The only way of causing this to happen was to find investors and develop it using a privately owned company.  A friend of mine, Peter Martyn, was a senior manager in a Tenix/Toll consortium that provided all 4th line logistic support to the Australian Defence Force.  Peter suggested that Tenix/Toll would be interested in exploring the use of this technology in the workplace.  This eventually resulted in Tenix/Toll engaging dPId to develop two systems; one called JTaR and the other called TIMS.  Both were intended to be demonstrators.  JTaR is still in use at the time of writing.  TIMS was successfully developed up to the point where it could be deployed but that deployment was prevented due to the failure of the Military Integrated Logistic Information System(MILIS), introduced by Defence in 2010.  MILIS was unable to produce stocksheets for over 6 months from the time of its introduction.

TIMS was successful from a number of points of view:

• TIMS demonstrated that the dPId identification numbering system was totally compatible with other forms of barcode identification and could work comfortably side by side with these other forms of ID.  If anything, the dPId system sat above these other systems, bringing them together as one cohesive method of item or process management.
• Importantly, the dPId method of identification was shown to require about 60% less effort than other unique ID systems that had a hierarchical segregation of the numbers, such as GS1.

To add to the difficulties of progressing this project, during this period of the project's development, Tenix/Toll sold the Defence contract to British Aerospace Engineering Systems Australia (BAESA) so there was a change of management to be contended with.  Both Tenix/Toll and BAESA were very difficult organisations to deal with and had many of the characteristics of the ADF; characteristics commonly found in large organisations where bureaucratic process suffocates all entrepreneurial spirit.

Investors

Investors for the development of technology were found in the form of Bligh Street Capital Partners, Bartter Enterprises and Mr Denis Lozina.  Bligh Street Capital Partners is headed by the Honorable Mr Neville Wran and his longtime business partner, Mr Alber Wong. Bartter Enterprises is owned by Mr Peter Bartter, former owner of Steggles Chickens.  Peter has subsequently become Chairman of the Board of Directors. 

Technology Development

With investment, dPId was formed and the technology development began.  Ivan Curtis developed an FPGA connected to a web-cam to demonstrate how it was possible for a reader to read a dPId barcode.  Experiments were performed with a laser etcher to create the barcode using this method and these experiments were successful.  During this time it became apparent that, within the funds available, it would not be possible to develop a fully-functional reader.  The prototype reader developed by Ivan required around 25 seconds to read a NU-ERA barcode.  Ivan was confident this time could be dramatically reduced with further development of an Application Specific Integrated Circuit and more powerful circuity.  It was hoped that funding for this further development could be obtained from Defence.  For the Tenix/Toll demonstrator system it was necessary to use DataMatrix barcodes to represent the dPId numbers.  To further demonstrate the use of this technology, a decision was made to develop a decryption dongle that could be attached to any computing device with a USB interface.  The decryption dongle had the ability to decrypt the number read by a conventional barcode reader and to check that the number was indeed a number issued from the dPId number registry.  In 15 Apr 2008, development on a stand-alone crypto module commenced and was over-seen by Mick Evans.  By August 2008, the module had been designed by GF Micro in the UK and constructed and tested by Ivan Curtis in Adelaide.

Submission of an Unsolicited Innovative Proposal(UIP) to Defence

In late 2010, an unsolicited innovative proposal was submitted to Defence at the suggestion of the Secretary of Defence.  The result was interesting in that it gives an insight as to why Defence has a problem with new technology, particularly the ADF which has a "small-military", "follow-the-leader" mentality.  Their reply stated, "While the 'NU-ERA 2D Bar Code System' is a great idea and provides a unique method for tracking, the problem lies in its interface with all other entities that Defence deals with.  It does not comply with the current ADF and allied standards, and for this reason its introduction into service would cause problems with existing systems.  It would require a world-wide change in current tracking methods, which would not be cost effective for Defence at (sic) present time."  Besides the fact that the latter assertion concerning the need for a worldwide change, was shown in the demonstrator system at the Defence National Supply and Distribution Centre to simply not be true, what this response was actually saying was that the system being proposed was novel, that is, innovative, and unless it is being used by others, the ADF would not consider it. So it would seem innovation is OK provided it's innovated somewhere else first!

Where to Next?

dPId needs to find further investment to develop its reader to the point where it can be used in the workplace.  This requires the building of an ASIC and the design of advanced, powerful, energy efficient circuitry to support the ASIC.  dPId also needs to raise finance to build a hand-held laser etcher as this will greatly facilitate the marking of items coming off production lines and in the workplace; including places such as the humble corner grocery store.