At a Tesco's supermarket in Cambridge, England, the shelves have begun to talk to their contents, and the contents are talking back. Soon, razors at a Wal-Mart store in Brockton, Massachusetts will begin to let staff know when they suspect theft. This spring, a group of firms will attempt to track, in real time, many thousands of goods as they travel from factory to supermarket shelf. Consultants tout cost savings and extra sales that could run into tens of billions of dollars a year.
The reason for the sudden buzz of excitement is a new, super cheap version of an old tracking technology called Radio Frequency Identification (RFID). RFID systems are made up of readers and 'smart tags'-microchips attached to antennas. When the tag nears a reader, it broadcasts the information contained in its chip. In the past four years, the cost of the cheapest tags has plunged, from $2 to 20 cents. In the next two to three years, prices are likely to fall to five cents or less. Already, RFID tags are made in their millions and used to track pets and livestock, parts in car factories and luggage at airports. Last month, Gillette announced that it had put in an order for half a billion smart tags, signalling the start of their adoption by the consumer-goods industry. If they catch on, smart tags will soon be made in their trillions and will replace the bar-code on the packaging of almost everything that consumer-goods giants such as Unilever make.
The inspiration behind the new, cheap tags is a partnership between academic researchers and business called the Auto-ID Centre, based in Cambridge, Massachusetts. Founded in 1999, the centre boasts 87 member companies, including the world's biggest retailers and consumer-goods firms. Traditional RFID tags carry all their information. That makes them big and costly-fine in small numbers, but expensive in the sorts of quantities that the consumer-goods industry might want. 
But nobody had got it,. Big technology firms such as Intel and Motorola thought it was impossible to build a tag costing a few cents. Traditional RFID makers, who grew up without the internet, did not understand the beauty of removing information from the tag and storing it centrally. So Messrs Sarma and Ashton did the work themselves, designing specifications for a new chip and inventing new software and network services to support their idea.Abandoning the likes of Intel, Mr Ashton and Mr Sarma turned instead to a handful of start-ups. One of them, called Matrics, says that it is now ready to start making the new tags. 