The data avalanche

LinkedIn +

Radio frequency identification (RFID) has become the predominant supply chain issue – in the public imagination, anyway. Seldom has an advance in supply chain management technology achieved so much mainstream media coverage.

Walmart’s commitment to ‘forcing’ suppliers into RFID, Gillette’s plans to combat theft of Mach 3 razor blades (allegedly the most shop-lifted product in history), proposals to link RFID to CCTV coverage, and the inevitable rise of powerful, allegedly consumer-led – trial lawyers are also consumers too, you know – pressure groups have all meant an unprecedented focus on what is still an emerging technology.

So where better to go for a more considered update on RFID than the relevant session at the ‘Logistics and Supply Chain Forum’ on the liner Oriana in October, somewhere in the English Channel?

At the Forum, it became rapidly clear that those worried about the aforementioned privacy and rights issues are getting well ahead of themselves; the questions are not unimportant but will only become issues once the industry has determined the business and logistics case for RFID.

Chairing the session, Jim Spittle of e-Centre pointed out that there is nothing particularly new about RFID but adoption has been surprisingly slow. There are issues already in play: understanding the technology; understanding how it could help the business; the issue of compulsion (Will my major buyer force me to use a technology that I am not comfortable with?). Walmart’s commitment to bringing its 100 major suppliers into line by 2005 and tens of thousands more soon after certainly colours that debate.

This is all about better control
Brian Weeks, MD of Integated Product Intelligence, admits there is ‘a lot of scepticism as to whether RFID will work in a real supply chain’ and like others he dismisses both technological and rights issues: ‘This is all about better control of supply chain assets, not about control over the people who legitimately or illegitimately pick those assets up.’

Technology, contrary to the pitch of many vendors, is not really an issue – the issue is commercial relations and what you actually do with this exponential increase in data. John Greaves of ePC suggested that the NAPCCI project (ie the Walmart initiative) had proved that there were no issues over frequency spectrum or the material you might be attaching an RFID chip too, that could not be solved. And he threatened: ‘Whether you are a manufacturer, distributor or retailer, NAPCCI is on target and on time in Arkansas. Some suppliers [to Walmart]risk commercial failure in 2005 and the other 86,000 in 2006 because there are not enough resources in the world to meet the demand for RFID.’

That is, perhaps, apocalyptic. Greaves, and other speakers, agreed that general application of RFID at the item level is unlikely for five or even ten years’ time although ‘you’d be surprised how many items already are tagged’.

But Greaves and others also condemned some of the current salesmanship. Perhaps no technology is ever fully developed but RFID is further on than you would know. There are, effectively, no frequency barriers world-wide: even Japan has made spectrum available, and in less than five years Ultra Wide Band technology will sweep the remaining inconveniences aside. (UWB is too complex to discuss here, but think of it like packet switching – you can essentially freeload on someone else’s frequency without inconveniencing them in the slightest). Further, we are seeing multi-protocol readers and, rather less effectively, multi frequency readers coming online. So even if we do not immediately achieve a global uniform standard, that doesn’t mean RFID can’t be done on a global basis.

In data, people are the problem
The parallel issue, and one which, given the tortuous history of bar code adoption, is likely to be more significant, is that of EPC – electronic product coding. There is already a model which, when refined, will have a huge impact on supply chain operations and data mining techniques – and enable ‘OHIO’ (zero human intervention operations). As Greaves claims: ‘in data, people are the problem’; if data could flow incorruptibly from supplier through manufacturer to point of sale and beyond, there is the possibility of a profound change in supply chain operations.

But the immediate problem with RFID and the barrier to effective implementation, is: what do you do with all this data? If you track a truck through in-cab telemetrics you have one lump of data. If you track each roll-cage, you have 40. If you track each tote, that’s perhaps another 20-fold increase. If you track each item, you could have increased the volume of available information 80,000 times. Most of this data is good but irrelevant.

Management by exception (ie only clocking the bits that don’t work) is vital, but that is more than a technological change. All managers want to report the 99 per cent of activities that meet the plan. To focus purely on the one per cent that don’t requires not a technological change, but a culture change. That is the true challenge of RFID.


Briefing points

  • Despite the media frenzy, privacy issues are not uppermost because item-level RFID is not going to become general for quite some time
  • That is not through any technical weakness: frequency, spectrum, material (ie devices that will work against bulk metal, or wood, or whatever) have largely been sorted.
  • Product coding standards will be vital, as anyone who remembers the beginnings of bar coding will be aware
  • Implementation will be patchy but obligatory, so find out what your suppliers and customers are planning
  • RFID, at the item level, is going to increase the volume of available data a thousand or million-fold. Most of this is merely confirming that things have gone well. How are you going to identify and act on the error and event messages? That is the number one issue.
Share this story: