Smart Weed Whacker Takes Guesswork Out of Spraying

In the not-too-distant future, tractors already decked out with air conditioning units, CD players and global positioning systems may be equipped with another handy gadget, according to Agricultural Research Service scientists and university cooperators.

The gadget in question? An optical sensor that detects weeds and prompts a herbicide sprayer to target them as the tractor drives up and down crop rows. The technology means less herbicide in the environment and lower costs for farmers.

Most farmers now apply herbicides uniformly over a field, but weed distribution is often patchy-particularly grass weeds in cereal crops. Kansas State University graduate student Ning Wang and KSU professor Naiqian Zhang, in collaboration with agricultural engineer Floyd Dowell of ARS' Grain Marketing and Production Research Center in Manhattan, Kan., and others have developed an optical weed sensor and spray control system that takes the guesswork out of herbicide application.

Using a near infrared spectrometer, the researchers studied the light absorption characteristics of weed stems and leaves. They identified five wavelengths that can be used to discriminate weeds from crops and soil and, with them, developed the optical sensor.

The current model requires the tractor operator to spend about 5 to 10 minutes "training" the sensor to spot different weeds. The operator does this by placing the sensor above a weed and typing in a "weed code" on a simple keypad. Once the weed sensor has enough statistical data, it prompts the operator to begin driving up and down the rows. If the sensor detects weeds, the herbicide applicator will spray them.

During laboratory tests, the sensor identified wheat, bare soil, and weeds with accuracies approaching 100 percent. Although these results are promising, the researchers believe it will be several more years before the system is ready to be commercialized.

ARS, the U.S. Department of Agriculture's chief scientific research agency, provided funding for this research, as did KSU, the National Science Foundation and the Advanced Manufacturing Institute.