Not long ago, drones on a farm meant male bees, essential for a healthy hive. But like catfish, cloud and viral, drone has taken on a new meaning in the age of high tech.

Drones — the small flying robot variety — are ushering in a new agricultural revolution, says information specialist Gerard Sylvester, editor of a new report on drones and farming by the Food and Agriculture Organization of the United Nations and the International Telecommunication Union. Estimates suggest the agricultural market for the machines will be in the billions of dollars in the coming years. As farmers try to adapt to climate change and deal with other challenges, Sylvester says, drones promise to help make the whole farming enterprise more efficient.

The drone technology “is already there,” says aerobiologist David Schmale at Virginia Tech in Blacksburg. But what’s rapidly advancing is the cargo the drones carry. “What you are seeing now is drones becoming the vehicle for really cool sensors,” he says.

When outfitted with cameras and other data-gathering devices, drones provide an “eye in the sky,” scouting for plant pests or dry spots in need of more attention. In some countries, drones are already regularly used to deliver fertilizers or pesticides; an estimated one in three bowls of rice served in Japanese homes is grown with the help of pesticide sprayed from an unmanned helicopter. Other tasks have yet to become mainstream. US farmers, including grape growers in California and New York, have been experimenting with drones to survey for areas of “low vigor” where water may be lacking or soil isn’t up to snuff.

While drones aspire to be all-seeing, there are still blind spots: Discerning a weed from a crop, for example, requires particular powers of perception. And earthly challenges remain, including how to analyze and interpret the reams of data gathered by the gliding gadgets, and how and with whom such data will be shared. Nonetheless, drones will probably soon become a standard piece of farm machinery. Here are some examples of the many potential jobs for these flying robots:

1. Crop assessment

Monitoring fields takes time and manpower. Drones can sweep through and take stock, inspecting for slow-growing plants that might need a hit of nitrogen or some other remedy. Sensors that measure particular wavelengths of light absorbed and reflected by plants can generate color contrast images that highlight problem areas in a field. Images generated from these data include NDVI (normalized difference vegetation index) maps, which have long been created using satellites and airplanes by calculating the ratio of the difference of near-infrared and visible light radiation. Newer sensors take advantage of additional wavelengths in the light spectrum and may add filters and other resolution-enhancing tricks. Not only can soil, crop and forest be told apart, but sick plants can also be spotted because stressed or dehydrated plants reflect light differently. Recent research shows that this spectral data can reveal crops that have been injured by the drifting pesticide dicamba, and can spot herbicide-resistant weeds growing between rows.

Two maps of a wheat field created with data gathered by a drone. One shows the field in infrared light; the other is an NDVI or normalized difference vegetation index of the same field.

Wheat watch: After a farmer in Ontario, Canada, learned army worms had infested some of his crop, researchers surveyed his fields with a sensor-equipped drone. The resulting maps highlighted areas for further investigation: The NDVI map (bottom) has a section infested by the army worms (B), next to a worm-free, healthy alfalfa field (A). A rocky outcrop is visible (D). An area where wheat stems are damaged (C) stands out as bright pink in the mosaic infrared color composite map (top).

CREDIT: C. ZHANG ET AL / PLOS ONE 2014

Differentiating a weed from a crop plant within a row is still a challenge. “Green is green,” says agricultural engineer Yanbo Huang of the US Department of Agriculture’s Crop Production Systems Research Unit, in Stoneville, Mississippi. He’s working on algorithms to determine such things as leaf shape and texture that, combined with imaging information, could tell cotton from crabgrass.

2. Counting cattle

Cattle ranchers with lots of land to cover are using drones to keep track of their livestock and survey where fences need fixing. When outfitted with high-definition thermal imagers and night-capable cameras, drones can also help survey for unwanted animals that might be ravaging a herd. (Such drones have also become a tool for tracking human poachers in places like India’s Kaziranga National Park, home of the one-horned rhino.)

Diagram shows the different absorption patterns of blue, green, red and near-infrared light characteristic of a healthy leaf, a dead leaf and a yellowed or stressed leaf.

Early efforts by NASA to monitor vegetation growing in the Great Plains via satellite led to the development of the normalized difference vegetation index (NDVI). This visual yardstick of plant greenery is possible because plant leaves absorb and reflect different wavelengths of light: chlorophyll in healthy leaves absorb visible light (some green is reflected, which we see) while reflecting near-infrared light. A yellow, stressed leaf and a dead leaf (as well rocks as soil) reflect and absorb these wavelengths differently. Drones equipped with sensors can gather this spectral data and create maps that show variability in crop health. Moderns tweaks — sensors for additional wavelengths of light or using filters or lasers, for example — and more sophisticated analyses offer improved resolution for farmers who want a quick evaluation of a field.

3. Monitoring for disease

Without close scrutiny, pathogens that wilt, wither and otherwise damage crops can escape detection and spread. While spectral imaging technologies can reveal yellowed plants within fields of green, Virginia Tech’s Schmale is using drones to ferret out high-flying pathogens before they even land. He has captured airborne spores of the fungus Fusarium graminearum, which can devastate wheat and corn, that have traveled kilometers and more. If a farmer knows of a pathogen outbreak in a nearby county, air sampling could provide an alert to its impending arrival. Federal and state agencies could also monitor for pathogens at a larger scale, allowing farmers to be at the ready before outbreaks occur.

4. Water watch

Many fields aren’t uniformly flat. Some sections may dry out faster than others or be missed by watering equipment. Spectral and thermal imaging can reveal dry spots where crops will wither. Imaging can also detect leaks in equipment and irrigation canals. What’s more, farmers can assess the topography of their land with airborne laser scanning technologies or software that stitches together thousands of high-quality aerial photos into 3D maps. These maps can identify water catchments, reveal the water-flow direction at the base of each tree in an orchard and identify other land features that influence both the health of the crop and where soil erosion might become an issue.

5. Mechanical pollinators

Actual bee drones don’t help with pollination, but the flying-robot sort may one day lend an assist to real bees. A New York–based start-up has developed pollen-dumping drones that have helped pollinate almond, cherry and apple orchards. The company reports that its drones can up pollination rates by 25 percent to 65 percent, though outside analyses verifying those numbers are yet to be done. But some fruit growers are optimistic that the drones could prove useful in orchards, especially if a cold snap keeps live bees hive-bound during the trees’ flowering window.