You are browsing the archive for Philippe Plagnol.

Open letter to GS1 : The Benefits and Potential of an Open GEPIR

- December 23, 2013 in Uncategorized

Dear GS1 and GS1 Member Organisations,

For over 40 years, GS1 has succeeded in creating a unique and standard identifier, what we know as the barcode, for the vast majority of commercialised products.  Through your 108 Member organisations operating across more than 150 countries, you have succeeded in simplifying trade in complex global markets. However, the rise of internet and communications technologies has meant that GTIN codes, the unique and standard identifier for products created to serve the needs of manufacturers and retailers, now has uses not imaginable 40 years ago.

In 1999, you created the Global Electronic Party Information Register (GEPIR), a system that permits consumers everywhere to verify the validity of GTIN codes by allowing users to discover the company hidden behind the barcode. Anybody, anywhere can use GEPIR to verify the company that registered the first 6-10 digits of the GTIN, the Global Company Prefix (GCP), a number unique to an individual company, and therefore trace products to their owner. For example, using GEPIR, using the GTIN code on my Coca-Cola bottle, 049000000450, I am able to discover that the GCP “0049000” is assigned to “Coca-Cola USA Operations”, a legal entity based in Atlanta, USA. In global markets, it is often difficult to know where individual products come from or who registered them, making the GEPIR system invaluable for conscientious consumers. We are aware that GEPIR also serves to help combat the use of illegal GTIN codes and recognise that GEPIR is an incredibly power tool that permits companies and consumers alike to trace the origins of the products they buy.

Despite being limited to 30 requests per day, GEPIR has always been free to access. Recently, however, we noticed that GS1 France changed their access policy, closing their repository to the public and creating “premium”  access for companies willing and able to pay (See Article). While we were still able to trace French products using the US GEPIR, we worry that closing the system to the public to offer premium access might become a trend, a trend we see as bad for the consumers and bad for GS1. We are proposing a different trend and urging GS1 to open up rather than close off.

For the last 40 years, the GTIN code has been used primarily for supply chain management purposes but in recent years we have seen barcodes used in myriad of smartphone applications. Changes in technological capacity has meant that the GTIN code can be used in ways unimaginable 40 years ago and while it was never intended to be used in mobile applications or as a communication channel for companies, it is evident that GTIN codes will continue to be used in new and innovative ways in the years to come. That being said, at the moment, use is limited by the closed nature of your current system.

For over a year now, we have been working to understand and build an “Open GEPIR”. In order to truly capitalise on the potential of GTIN codes, an open product database is necessary and building a reliable database requires quality data. GEPIR is essential both to verify the quality of our data and in linking individual products to the company that produced them. GEPIR is the key to unlocking the potential of Open Product Data and creating genuine traceability and transparency in supply chains. We have made a good start (see our current progress here) but only you, GS1 and your member organisations, can provide comprehensive data dumps and real time synchronisations of data-flows.

We urge you to join us in enabling innovation in the product space and capitalising on the potential of the system you have built. We ask that you make open data part of your 2014 strategy!

Seasons Greetings,

Philippe Plagnol and Open Product Data Working Group

Brand Standard Identifier Number (BSIN): Open Product Data Working Group Launches a Brand Repository!

- December 20, 2013 in Uncategorized


In order to build a meaningful database of all the products in the world, it is important that we are able to link essential attributes of the product to the product itself. One obvious attribute that we have wanted to link is the product’s brand. Unfortunately, we discovered that assigning a brand to a product is not as easy as we anticipated. Not only is there no standard identifier for brands, there is also not always a clear distinctions between a brand, a product line, a company etc. Nevertheless, a product’s brand is incredibly powerful in influencing purchasing decisions and we felt that it was important to develop a way of managing and assigning brands to products.

The working group had already built a brand repository containing upwards 4,000 brands but this repository was highly integrated with our open product database and our system of assigning brands to products was not scalable.  To correct for this we have developed a separate brand management platform and created a six digit alphanumeric code, the Brand Standard Identifier Number (BSIN), unique to individual brands. The brands already in our product database have now been assigned a unique BSIN code. The brand repository is available under an MIT licence via our API or by bulk download.  We can easily foresee the use of BSIN by all the product related applications and its use in the APIs of other system.

Ultimately, it will be possible to group brands by brand owners (i.e. all the brands belonging to Unilever or Pepsico) using OpenCorporates, for example, or link brands to resources such as wikipedia, and find the history of the brand using organisational registries like the INPI in France. 4,000 brands is only just the beginning. We will continue to add brands to our repository and users can recommend brands to add to the repository as well.

If you would like to be involved in the management of this repository (no technical skills required) or to participate in the development of the brand repository platform (Python / Django / Postgre), introduce yourself on the mailing list or find us on Github. We have a long way to go before we can truly link every product to its respective brand and owner but being able to identify and assign brands is an important first step.