Robotics and Autonomous Equipment
The United States definitely doesn’t lead the world in autonomous agricultural equipment development or use. Although the U.S. military probably is the leader in autonomous operating systems, and there really isn’t much of a reason why this precision technology cannot be transferred from military to agriculture except for cost of sensors and communications systems.
“As a company, we recognized that you cannot take guys who have operated for the military or the government and expect them to be successful in operating a for-profit business in a sector like agriculture,” said Young Kim, general manager for BOSH Precision Agriculture.
The government works with very different metrics of success than a for-profit business. “The military cares less about economics and more about outcome,” said Kim.
But Kim is going against his own rule to a degree switching from supplying to the military to an entrepreneur business philosophy of offering service to the agricultural industry at the yet-to-be approved level of unmanned autonomous aerial systems.
Kim provided his concept of making a positive economic impact in agriculture with an autonomous helicopter and spraying service during the Ag Connect Expo and Summit in Kansas City at the end of January.
OTHER POINTS OF VIEW
Helicopter spraying with unmanned aerial systems could replace manned helicopters if government regulations would allow. Also, providing insight into the autonomous/robotic vehicles potential and state of the industry was Mel Torrie, president and chief executive officer of Autonomous Solutions, and Stewart Moorehead, manager of robotic systems with John Deere Technology Innovation Center.
Torrie says a lot of robotic/autonomous vehicle technology can go from research to field for agriculture and mining. Large-scale equipment is appropriate for both industries, but mining operations are done in a much more controlled environment than agriculture. Autonomous trucks for hauling mine diggings are already in use.
String technology has been used by the military for a caravan of trucks in the Middle East with the lead truck being the only one human driven. The technology is already in field demonstration by ag equipment manufacturers for offset tillage and harvest operations by two or more tractors or combines. Torrie sees it as the method of moving autonomous equipment from one field to another.
“Our approach is to develop incremental autonomy with the OEMs (original equipment manufacturers) and then incubate full autonomy with large end users,” the company executive said.
Moorehead explained that John Deere has already developed a lot of automated manned equipment systems. “The iTEC Pro is automating everything on the tractor,” he said. But what it doesn’t include is “perception safeguarding” which is necessary for autonomous machinery operations.
Moorehead described a field research project being conducted in an orange grove in Florida. This autonomous unit of tractor-pulled mower and sprayer system has proven itself to be efficient and worthwhile for an orchard or other types of specialty crop production.
ORANGE GROVE RESEARCH
“The system has the perception for safeguarding—the ability for the vehicle to see what is in its environment and understand what it is able to drive through, such as four- or five-foot tall grass, and what it is not able to drive through—four- or five-foot tall people, utility vehicles or a pickup truck. The system we are using makes use of multiple sensors,” he said. “We combine things such as laser scanners, color cameras and even a radar system for identification.”
This supervised autonomy mower and sprayer unit stops when it needs help to determine if an obstruction is a concern or a false alarm problem. One operator is available for being messaged via smartphone or a computer. The simplicity of the system in this controlled orchard grove is allowing one operator/supervisor to oversee two autonomous units. As Moorehead noted, in the last 18 months the supervisor operator has been contacted on average every 20 minutes, which means one operator could oversee up to 10 of these units in a controlled environment.
He explained how the supervised autonomous units with “optimal mapping plans” are loaded with a computer logarithm for determining the shortest path so that they are averaging 10 percent shorter distance than a manual-driven unit.
Additionally, even more importantly, Moorehead said, the autonomous units have spent a greater percentage of the day operating at maximum speeds than a manually driven unit, which has accounted for about a 30 percent increase in productivity per tractor.
An orchard is a difficult environment to test under because of the misalignment of some trees and limbs, tall grass hiding potential obstacles and tight quarters of little unused room. But the best aspect of an orchard is that it is a controlled access area where the autonomous unit does not have unauthorized people or vehicles wandering through the environment.
Moorehead said engineers have to approach solving problems through autonomy and robotics by completely understanding the way that the customer works and how a system can fit into specific operations.
“All the things on this vehicle (orchard unit) can be used to work in corn and soybeans on an 8000 series tractor or a 9000 series tractor,” he said. “The one thing that people don’t understand too much about robotics is how many of these technologies, and many of the strategies, are very ‘generalizable’ to the full range of agricultural operations.”
CHALLENGES STILL AHEAD
Torrie noted there are challenges for autonomous equipment being built for growers by non-major OEMs. When a unit needs service, repairs and parts, they need to be just as available for the autonomous system as for conventional equipment. He thinks sales, distribution and support are huge challenges for a non-mega equipment manufacturer.
John Deere has a proven results project to replace the driver in mowing and spraying operations in an orchard with an autonomous unit. Some people see the autonomous and robotic endeavors as just attempting to replace the driver. Taking the driver cab and control pod off of a tractor could save up to 30 percent out of a tractor’s manufacturing cost, he contends.
Restructuring the size of equipment might be one of the most far-reaching changes. He suggested four small autonomous units might replace one large conventionally manned unit. The breakdown of a small unit leaves three still running, but the breakdown of the large unit brings everything to a halt.
Safety is one of the biggest issues, and types of sensors to do its specific agricultural task plus identify obstructions and other vehicles are key. Those sensors exist; it is just that these sensors are quite costly. But Torrie credits Google and the automotive industry in making headway so that supervised autonomy will be accepted as safe and for the production of lower-cost, mass-produced sensors needed in agriculture. Google is pushing governments to change regulations outlawing much of the autonomous industry.
Change in regulations is definitely needed in allowing Kim’s ideas and autonomous technology service to become widespread. Currently, Federal Aviation Administration regulations don’t allow large-scale autonomous aerial vehicles to fly except under the supervision of university researchers.
ROW CROP HELICOPTER SERVICE
But of course, Kim sees his helicopter service as being more efficient and environmentally friendly as well as more economical for helping all types of growers earn a profit.
“Agriculture has the desired market dynamics” for an entrepreneur to enter and make money, he said. The size of the market is great for an entrepreneur. The two components of BOSH in agriculture is providing UAS imaging from multispectral sensors and crop spraying from an unmanned helicopter. Kim expects to determine where a crop is unhealthy and then help do something about it.
Kim has several reasons that an unmanned helicopter fits into large field operations. His list includes better quality of spraying with down force from the helicopter rotor, elimination of wheel crush effect of the crop, eliminating cross contamination from one field to the next, safety of a no human low-altitude fast maneuvering helicopter, ability to program optimum height for controlled drift, night application when many crop protection products work best and allows a much longer application window per day and finally more efficient flying than a manned unit as the unmanned helicopter flies forward and backward without turning around to go down rows.
U.S. regulations are a hindrance that other countries don’t put on UASs. “There are more than 2,700 of these (small) crop sprayers flying today in Japan. There are more than 15,000 licensed operators in Japan just doing crop spraying, primarily spraying rice in small plots,” Kim said. The unmanned helicopters are typically about 360 pounds and eight feet long.
Kim foresees full-sized autonomous helicopters flying everywhere to service growers’ needs, and he wants regulations to allow this as soon as possible. It’s not the technology that is limiting progress.