Over the next decade, wireless sensor technology will change how data is collected from physical objects. For example, a new heart sensor developed by Corventis (corventis.com) — one that the patient wears during the day — reports anomalies to a doctor, who can trace historic sensor readings; IBM is developing bridge sensors that monitor traffic levels and can predict structural integrity issues. Yet one of the major challenges is in storing the data, analysing it, and making decisions about the analysis. Collecting the data is one step; knowing what to do with it is another. For the past few years, Teradata (teradata.com) has experimented with wireless sensor technology in Californian vineyards, storing data from thousands of vines in order to help vintners know the exact time the grapes should be harvested. The sensor data is extensive: it involves temperature readings, moisture levels, and sugar content. Not only are there millions of data points for one vineyard, collected from each vine, but the analytics show the inter-relationships of the data. For example, vintners can see when moisture levels and sugar content are perfectly aligned for the best wine."We have a massive distributed network of sensors we can store," says Stephen Brobst, the chief technology officer at TeraData. "In the past, the vintners would have to experiment with grape production and go by taste. Now, we can analyse with more granularity — reading data from each vine." Brobst says current sensors are about the size of a finger nail, but will be smaller than a grain of sand within a few years. Researchers at Cornell University are also experimenting with micro-sensors for vineyards. In one study, researchers embedded a 2mm-wide chip into the trunk of a vine. As the vine grows around the sensor, the vintners can see historical data about the water levels and other environmental factors, increasing the accuracy of data collection. This fall, Cornell will create a prototype of the embedded sensors at the Cornell Nanofabrication Facility in New York. Brobst says wireless sensors are already used in the shipping industry — for example, reading the number of TVs in a truck without having to peek inside — and will eventually be used to monitor human behaviour.