Using what little information is available about land and expected weather conditions and water rights, canal managers must often decide how much water to divert several days before it reaches its destination.  Crops can easily wither in that time.

In an industry dealing with such a valuable resource and operating on such a tight margin, flipping the switch one way or the other can have significant consequences.

Researchers at Utah State University’s Utah Water Research Laboratory and Center for Self-Organizing Intelligent Systems are using small airplanes to remove some of the guesswork behind these critical water decisions. 

“Unmanned autonomous vehicles, or UAVs, can help provide objective information for more efficient operation of irrigation water delivery systems—on the cheap,” said Mac McKee, director of the UWRL and member of the research team, which also includes YangQuan Chen, Wei Ren, Hui-Fang Dou, and several engineering students.

Flying above irrigated fields, canals and bordering areas, the UAVs take pictures below that indicate the moisture level of the land.  Dry lands need more water funneled to them, and overly-wet lands could indicate excess canal seepage.

“By stitching these high resolution photos with existing satellite imagery, we can create an information-rich map that pinpoints areas that need more water and areas where the causes of canal seepage must be repaired,” said McKee.

The military has used UAVs for years in reconnaissance, logistics, and combat and has often contracted with USU’s Space Dynamics Laboratory for their development.

“We saw this as a great opportunity to collaborate with another unit on campus, but we had to do our cutting-edge research on a shoestring,” said McKee.  “We saw what SDL was doing with their UAVs, but, whereas they were building theirs on beefy military contracts, we had to produce ours at costs that are two orders of magnitude cheaper.”

Instead of creating a UAV for $150,000 on a military budget, said McKee, the UWRL research team had to do something similar for under $1,000 per unit on a water system budget.

“This project is important for water managers because every irrigation company on the planet is on the edge of bankruptcy and doesn’t have a lot of money to throw around,” said McKee.  “Our job was not just to create a solution but to design one that cost less than the value of the information it generates.”

To work on such a limited budget, the research team had to start from scratch.

“We rebuilt the whole nine yards,” said McKee.  “The computer, CPU, memory, GPS, avionics, camera—they all had to be redone in a way that integrates with the mission we envision.”

What the team came up with was a fleet of hobby-shop looking airplanes.  Decorated in red, green, and blue, the UAVs look recreational, but they pack a scientific punch, equipped with a computer system allowing them to fly short missions without human contact.

“In March [2007], we didn’t even have an airplane model to try,” said McKee.  “Now, we have several models to which we regularly assign missions and throw in the air, and they complete their missions and come back in one piece. Our task now is to regularize the process and get humans completely out of the loop, especially with respect to data processing and analysis.”

In the past few months, the team has cleared some of its highest hurdles in getting the UAVs production-ready.  Although the UAVs could fly, land, and take pictures in between, there was the issue of dealing with the data once it had been collected.
“The airplane can take the photos, but it’s a whole other story to accurately determine where the picture was taken,” said McKee.  “We needed some type of geo-referencing to track the pictures’ location.”

The team believes that by using a second airplane flying at lower elevation, the UAV could match photos to the land below.  With the second airplane in the shot, the data will have a specific point of reference to work from.  This is one of the ideas that will be tested this summer.

With software developed by the URWL, the photos are stitched together, interlaced with satellite images, and viewable maps are created with detailed information about soil moisture and crop cover.

“In the future, farmers will be able to view the photos online, dial into their land plots and look at moisture forecasts,” said McKee.“ From those photos, they’ll be able to better decide whether to irrigate.  Information generated from the same data will be available to canal and reservoir operators—including more accurate forecasts of future water demands—so that they can more efficiently operate the water delivery systems.”

In the meantime, clients are lining up to get their own UAV system.

“The UAV-mounted camera has a hundred possible applications, and we just happened to choose the hardest one,” said McKee. “After we get the kinks out of this project, we can quickly adapt it to measure some aspects of water quality, water velocity, water temperature and lake circulation. One client, for example, would like to use the UAV to track the movement of fish in large rivers.  We are also currently negotiating with state and federal agencies in Texas and California for contracts involving agricultural and environmental applications of our UAVs.”

Later this year, the UAVs will be out the door, and McKee and his team will have developed yet one more solution to improve Utah’s water management system.