Doing a little light reading and getting a big laugh from the GAO's September report to Congress "FOOD SAFETY: Agencies need to address gaps in enforcement and collaboration to enhance safety of imported food." This report has plenty of points to comment over and today I'll start with the easiest. One of the four gaps identified is "CBP and FDA do not provide unique identification numbers to firms." I have extensive blogs and publicly commented on CBP 10 + 2 rule about this very issue. Once again the solution is a GS1 data structure called GLN, Global Location Number.
Let's start with CPB's arcane MID identifier. It was developed during a time when there was a need to harmonize the world's manufacturer identifiers and no practical way of doing it. CBP developed a rule on how a broker is to code a manufacturer and their address so that there might be some consistency in import records. The system does not work well. Different brokers apply the rule differently and come up with unique MID's for the same manufacturer. With 10 + 2 there is a crying need for unique identifiers for the seller, manufacturer and consolidator which have not been developed. On the 7501 entry summary the MID is the seller, who may or may not be the manufacturer. Until CBP gets the MID issue resolved there will never be a good communication between CBP and the other agencies, FDA included.
The basis of GS1 is the UPC code seen on everyday products in the grocery store with a scan code broken into two parts. The first set of numbers identifies the manufacturer and the second set the product. Combining the two the store knows where the product came from and what it is. From the first subset identifying the manufacturer GS1 developed the thirteen digit GLN code using the manufacturer subset of seven digits leaving five digits to identify unique locations. The last digit is a check digit. Five thousand factorial is a huge number. The scale of this system could handle a million Walmart's so CBP shouldn't get so uppity about their numbers being too big for commercial solutions to handle. I believe Walmart's data intake makes CBP's look like a pimple so a little reaching out for best practices would be appropriate here. GS1 is a global non profit which should be reassuring to CBP on two counts. One it is unbiased and second its Global. It's a solution that reduces the Tower of Babel to one language with agencies in the United States communicating among themselves as well as with foreign ones! It is exactly what the GAO reports asks for and it is the solution to make all these internally developed applications able to talk to each other.
How many times have we heard that Government needs better IT and more money to throw at it. Until an adoption of GS1 data structures is initiated every nickel spent developing unique CBP and FDA codes are a waste. The 2002 Bio terrorism law instructed FDA to initiate a prior notice system where every manufacturing facility register. No thought was ever given to the GS1's GLN data structure set as an ultimate conclusion. Now FDA dutifully collects data about imports through CBP using their own arcane language and blames CBP for not helping it understand what it has. Frankly I consider FDA's prior notice one big black hole of a data repository that is so unwieldy and incomphensible that no decision ever gets made from it. Before FDA asks for more regulation and money it should make sense of what it is has. In that bio terrorism data resides the tracing, tracking, holding, reviewing and fines issuing authority that FDA seems not able to handle. GAO's conclusion that FDA needs more congressional authority is incorrect because the FDA first needs to learn how to use what it has.
"We recently identified information technology management concerns that might hinder the rollout of FDA modernization projects such as PREDICT" says the September 2009 GOA report on Food Safety on pag 37. PREDICT has all the characteristics of a failed fragmented IT inititiative because no thought has been given to the data structure and the enterprise architecture that may be used by all the agencies in a coordinated assembly of data. I keep harping back to two key points, the structure already exists and the key player in commerce, Walmart, is already taking it on. Of course I am talking about GS1.org and it's data pool synching procecedure called 1SYNC.
Government agencies take pride in fact that the scale of their data needs is huge, much bigger than what commercial enterprise is capable of delivering. Well I beg to differ. Walmart's scale is on the same order of magnitude as government's in every measure: geographic, departmental and depth. According to the GAO the FDA has spent $9 million on PREDICT and will require millions more in next few years. It is certain th other agencies have their own inhouse projects creating more silos as we speak. I propose that GAO take on the language of GS1 and insist that every agency convert to that data set to unifiy all of them into one language. A language that GAO could use in datamining searches to answer questions that Congress asks. A language that lets CBP FDA and USDA among others speak to each other clearly while yet allowing each agency be as specific as required with the richness and depth of the data. This would be a languange that gives industry the possibility to publish their specifications to every government agency required to be advised of a specific import. Come on GAO contact education@1SYNC.org and ask what they can do for you. For no money at all they will give you code structures that would allow FDA to talk to USDA and Tabacco, Alcohol and Firearms to CBP. Let them show you how each agency can design the most comphrensive forms that their little silo hearts could ever think of with every little specific field filled in from the outside when commercial participants publish to the data pool's.
Had a long conversation with a company that was able to institute a thorough tracking and tracing system with 7-11 in Japan. It was a costly and process changing experience but now that the cookie cutter has been built they are eager to find more customers who want something similar. I know how they feel.