ObjectStore keeps tabs on RFID data

This week ObjectStore, a provider of technology designed to manage massive amounts of data in real time, will unveil RFID Accelerator. The company aims to give to supply chain capabilities similar to those used by financial institutions for executing programmatic trades on the stock exchange. 

ObjectStore’s products use what is called Complex Event Processing, a method that queries by time, space, and identification rather than the more typical relational database type of queries, sometimes called an associative lookup.

RFID Accelerator includes ConnecTera RFTag Aware middleware to integrate real-time RFID data into existing applications and a query adapter for querying event history using SOAP. Both the queries and the results are represented as XML-formatted specifications.

RFID Accelerator uses both event data caching and reference data caching for context and to achieve its high speed throughput.

The company’s first offering, ObjectStore Trading Accelerator, monitors 750,000 stocks in real time and can log and analyze 50,000 trades per second, and execute a trade against a complex, predefined set of trading rules within two seconds.

Although RFID data may not be as rapid-fire as stock trades, analysts predict that within a few years RFID tags will generate terabytes of data daily. The notion behind real time monitoring of supply chain data is to move from a passive model using data warehouses, where data is analyzed after the fact, to an action- oriented model that can execute tactical up-to-the-minute decisions, said Richard Winter, founder of Winter, a consultancy specializing in large database technology.

DABAC (Data Bases à la Carte), a German software company, is trialing RFID Accelerator in a solution it created for a German forestry company Cambium-Forstbetriebe.

In order to stop a problem with inventory shrinkage of 10 percent to 15 percent between the time the trees are harvested as far away as Asia to the time they reach the German mills, a nail-shaped RFID tag will be inserted into every tree and tracked from the moment it leaves the forest until it reaches its final destination.

Scanning as many as 1,700 tags a second, the expected number of logs to arrive at any point can be compared with what was actually sent.