Drones Bringing Big Data to Agriculture

Collecting meaningful data for agricultural purposes has historically been a major challenge, mainly due to the tremendous land areas involved. The average farm in the United States has 441 acres of land, which is a little less than 3x the size of the core UC Berkeley campus encompassing 178 acres. Moreover, 41% of land in the United States is used as farmland, which amounts to a total of 938.28 million acres of land that  are used to grow a multiplicity of crops. The sheer size of the landmass of modern farms creates a great challenges in gathering meaningful data to improve the agricultural process. Irrigation data and soil composition data are just two examples of valuable information that would help farmers make smarter decisions about how to use their land more effectively. Traditionally, the only methods to obtain significant data for farms has been satellite imaging or flying over the land with a camera-equipped airplane. Both of these methods are expensive and often unaffordable for a substantial amount of farmers. Hiring a pilot to survey a farm can cost ~$1,000 per hour and satellite imaging can sometimes yield inaccurate results for certain weather conditions, such as thick cloud covers.

Satellite Imaging To Spot Erosion of Soil – Source: BBC

The arrival of low-cost unmanned aerial vehicles (UAVs), commonly known as drones, may soon change the game for data driven farming, since drones provide a low-cost alternative for farmers to survey their land for many different types of metrics. While hiring a pilot can cost ~$1,000 per hour, a survey drone can be bought for about the same price equipped with cameras, sensors, GPS and sophisticated autopilot software. The electronics and software of the drone enable it to collect the data farmers are looking for without needing a pilot. The cameras record pictures of the farm from above and can point the farmer to areas that may require further irrigation or crops that have been infected with pests. Furthermore, if the drone is equipped with the right type of sensors, it can capture images outside the visual spectrum, which can give information about soil content and other valuable metrics.


Top: a drone flying over a farm – Bottom: an infrared image showing chlorophyll levels – Source: MIT Technology Review

The price of small UAVs has been decreasing significantly, as the electronics and software components that make up the drone have gotten cheaper and cheaper. UAV designers now have access to open-source software for navigation and autopilot features that allow the drone to complete without user assistance. These autonomous features allow for farmers to have multiple drones survey larger amounts of land simultaneously or to program a drone to analyze different areas at times of their choosing.

Comparing Aerial Images of A Plane and a UAV – Source: BBC

The arrival of cheap, functional drones to the agricultural industry may be the key to bringing smart-grid type data analytics to a critical resource sector. If successful in the agriculture sector, drones may be used as data collection vehicles for other resources purposes, such as water flow or general land use. A promising start-ups that are already providing UAVs for farmers include HoneyComb, an Oregon based company that won the Cleantech Forum 2014 pitching contest, whose primary focus is to make an optimized UAV for farming purposes.