Organic is great, but in many parts of the world, farmers can’t sit around while pests eat their crops. Like it or not, insecticides are an important part of global agriculture. But what if we could at least use a lot less of it?

When you see farmers dousing their fields in insecticides, you might think that there are bugs in every corner of the field. But, according to Eamonn Keogh, a professor at UC Riverside, that’s often not the case. Unwanted insects sit in clusters, he says. And if you know where they are, you can target them, without spraying so indiscriminately. That means less pesticides for the rest of us.

That’s the idea of a project being developed by Keogh and his colleagues, which aims to give farmers more detailed information of where exactly insects are, and consequently where to deploy pesticides. He hopes the project will help cut pollution, reduce costs for farmers, and possibly alleviate the problem of insects becoming resistant to certain chemicals.

The idea is to put multiple sensors all over a farm, gathering information on insect locations hour-to-hour. By feeding that data into a central server, and combining it with information like weather forecasting, the system can then generate "heat-maps" for the farmer showing what areas need to be sprayed, and which areas can be left alone.

"I realized that if you knew, magically, where all the insects were, you could do very magical things," says Keogh. "Right now, if you want use pesticide on a citrus or avocado field, you typically have to go up in a plane or helicopter and spray the entire field. [With our device], you could spray a tiny bit in a localized spaces, and get the same effect."

Keogh compares spraying a whole field to a doctor treating a burnt fingertip with an all-over-body cream--overkill, in other words (though he does concede that some infestations require more drastic responses).

The device contains a laser pointed at a light-sensitive diode, or switch. When a bug crosses the beam, it generates a signal, which can be amplified, and analyzed, to suggest which animals are where.

Currently, farmers often use a sticky trap, which means they have to visit fields in person--a time-waster--and become semi-expert entomologists. Keogh says farmers find it hard to tell the difference between bugs that are going to do harm, and those that are actually beneficial, for example for pollination.

The project, which UC Riverside researchers are developing with a private company, has just won a prize from the Vodafone Americas Foundation. The sensors will probably be linked via a cell-phone network, though the exact details are still being worked out. The device is yet to be tested in field conditions. And the researchers cannot yet to distinguish between all insect species--some bugs are good to have around.

Keogh says he would like to sell the device to farmers in the developed world, as a way of subsidizing deployment in poorer countries. "We’re going to sell this to American producers, who’ll pay a premium to get beautiful, spotless apples, and then use the profit to help the developing world," he says.

The cost of the sensor has already fallen from $400 to about $25. But Keogh thinks a mass-market manufacturer, like Sony, could make them at about $2. The important thing, he says, is that the sensors are affordable and not worth stealing for parts.

The bug sensor is just an idea Keogh is pursuing as part of his plan to bring computer know-how to insect-related problem. He thinks "computational entomology" could also help with malaria research, and other agricultural pests.