Software Needs to Get Smarter

For RFID systems truly to deliver on their promise, software will not just need to be able to filter data, but to use it in entirely new ways, as well.e
Published: November 27, 2006

Mark Roberti, a British Web site focused on information-technology issues, published a recent article claiming CIOs fear their companies are drowning in data (see CIO Jury: Data Overload Drowning Businesses. The situation sounds serious and will only get worse as radio frequency identification systems exponentially expand the amount of data companies collect.

“Data management—and that includes the storage technology behind it—and knowledge management are going to be key technologies in the battle to remain operationally compliant [with regulations] and commercially competitive,” Gavin Whatrup, group IT director at marketing-services company Creston, is quoted as saying in the article.

Personally, I think this idea of data management and knowledge management is old school. Today’s IT systems are based on historical data. You collect information on what happened in your business, you analyze it with software and you come up with a plan for what to do next using software, RFID, sensor technologies, wireless networks and so forth.

It’s all about real-time data. Just doing a better job of managing more information is not going to give you a competitive advantage—using real-time data in real time will provide that advantage. RFID systems are going to force a major change in what software does.

I was speaking to an end user recently, who said RFID interrogators are more than capable of reading tags. What they now need is better logic, and to interpret the reads in real time. So, let’s say you’ve just picked up a pallet of goods, and the interrogator on the forklift truck reads the tag continually for about 60 seconds, then no longer reads it. The interrogator should interpret all of that real-time information and conclude that a put-away operation has occurred and send information to the inventory-management system.

Of course, that’s a very basic example. Here’s a more advanced example: A forklift truck backs up into a pallet of Tide detergent bound for a Wal-Mart or Metro store. The forklift truck moves the pallet to the damaged goods area of the warehouse. An RFID interrogator picks up the pallet ID, and IT systems record that the pallet was damaged. Today, it would be up to the warehouse manager to determine how to make sure the customer got the shipment. He or she might divert a pallet from a smaller customer to keep a larger customer happy, but that might not be the best solution.

What if software could dynamically analyze all the pallets of Tide within an eight-hour drive of Wal-Mart? Using RFID to determine what’s on the trucks, and GPS to determine where the trucks are in real time, then examining purchase orders and delivery dates, the software could reroute the trucks to ensure that all pallets of Tide are delivered on time, avoiding a possible out-of-stock situation.

Companies live or die on information. The ability to interpret the meaning of data and act on it swiftly will be crucial. Human beings absorb a wealth of information every second—sights, sounds and smells, as well as abstract information, such as words or numbers on a page—and yet we rarely feel as if we are being overloaded (except when it comes to e-mail). Software systems need to be more like our brains and less like our pocket calculators.