Automation is now the norm for the manufacturing industry, but how has the dairy farming industry overcome integrating robots with livestock?
The forces of innovation currently driving smart dairy farming fall into three main areas. The first is robotic milking, the second is automated feeding, and the third is concerned with the overall subject of quality: quality of product and quality of life for the animals.
Robotic milking is not a novel concept, it has been prevalent in the industry since the 1990s, but the real innovation is the intelligence behind it. Reducing labour costs through automation and intelligent software is the primary motivation for many industries, and large-scale farming is no different.
"It's so difficult to find good personnel who want to work in such a tough environment," says Peter Manneke, solution manager at DeLaval, a farming equipment manufacturer that specialises in making dairy farming smarter. "It involves long hours, getting up early and working until very late. It's an environment which is cold in the winter and too hot in the summer. It doesn't smell as nice as you'd expect in comparison to working in an air-conditioned office and it becomes harder and harder to recruit staff."
A changing industry
The perception of the agriculture industry as traditional and resistant to change is an inaccurate one. Although small-scale and artisan farming has seen a revival in recent years, many traditional farmers simply want to improve their productivity and yield, much like any other supply chain, and consider automation to be the way.
Manneke says negative views of smart or large-scale farming generally do not come from farmers, but from the public, and that in recent years there has been a noted rise in traditional farms integrating smart technology into their facilities. "There is a misconception from the public of what smart farming is really about. But trying to convince farmers to change to automated is like preaching to the converted."
In the mid 1980s, companies began pioneering designs for automatic milking. By the late 1990s many prototypes had hit the market and were being integrated into larger farms. Unlike early industrial robot models, the challenge of designing for this market was the very product it would be interacting with: livestock. Whereas robots in the industrial sector tend to interact with set points of contact on static components made of metal or plastic, a cow is never immobile and the udder shape from cow to cow is highly variable.
"We overcame this by using smart electronics, smart cameras and lasers," says Manneke. "The cow enters the stall and we use intelligent bespoke software to identify it so the machine can prepare for milking." Identification is done via an RFID ear-tag, which has become a standard for the industry, while the electro-polished, hydraulic robotic arm attaches to each teat. Teat visualisation systems feature an optical camera coupled with dual lasers to ensure teat localisation is carried out accurately and without causing distress to the animal. The robotic arm, designed specifically for the dairy market, allows for udder irregularities and movement of the animal.
For those who think large-scale farming can't possibly be a beneficial or comfortable experience for the animal due to pressures of increased milking and a lack of human element, Manneke says cows actually prefer the presence of a robotic milker to a human being.
"I have worked with cows for many years, and you think they know, trust and identify with a farmer. But a machine always looks the same, always sounds the same and always smells the same. It's never in a bad mood, it's never drunk and it's never sick. They are herd animals, but only when they are stressed, so a machine for a cow is not a threat, it is the complete opposite."
Robotic milking has recently celebrated its second decade in dairy farming, and the innovation behind this technology is now mostly confined to improving sensing and the engineering of the robotic grabbers.
The real emerging technologies in dairy farming lie in the software that interfaces into it. This is primarily in herd management systems, which store and 'analyse herd data for the farmer, creating bespoke profiles for each cow which are fed back to the automated milking, feeding and cleaning systems, often wirelessly. Farmers are now learning to harness emerging decision-making software and automation technologies to improve milk yield, quality and profit.
Smart farming places the health of the cow above every other factor because the healthier the cow is, the higher the quality of the milk, and the more frequent the yield.
Herd management systems use automated measurements to analyse milk samples and the temperature and weight of the cow to determine a variety of factors that will control the feeding and milking rate, or simply alert the farmer if there is a problem. This is particularly smart considering an average herd consists of 500 animals that must be monitored individually.
Through smart technology a farmer can now detect disease, whether a cow is in season, whether she is pregnant, the quantity of milk produced, the conductivity of the milk and the fat or protein content.This is done through a variety of sensing and testing. Representative milk samples are taken from specific milking points and sent to the analysis unit where the system selects which aspects should be measured in which cows, and during which milking session a particular cow should be monitored.
The milk is then checked for enzyme, ketone and progesterone levels, as well as other chemicals prevalent in the milk, to check for diseases such as mastitis, the biggest cost to the dairy industry. When a red flag is raised the herd management system can recommend a treatment or detect whether a particular cow has a cyst, with all data monitored remotely by a vet.
Cross-contamination can have a catastrophic effect on the food and beverage industry, so each new product that comes to market is subject to rigorous testing, both before use and over the lifecycle of its service.
DeLaval partnered with National Instruments to create a production test system for automated milking points using LabVIEW, CompactDAQ and NI TestStand, equipping the automated rotary system with a real-time electronic control to regulate milking. The milking point is a distributed system and communicates with a control unit through a controller area network interface. Testing includes sampling flow sensors connected to a specific milking point.
A vacuum regulation module, designed with flow sensors, pneumatic valves and regulators, ensures control of vacuum flow to the test object. To avoid contamination of the milk, seals and defects such as cracking in the milking point are detected, measured and logged by the system as a HTML file. Characteristics of the milking are also tested to ensure all equipment complies with government regulations.
Automated feeding provides better overall control of milk production, analysis of feeding effectiveness, and adjustment of rations in response to milk yield. The systems know what and when to feed each group of animals, saving huge amounts in labour costs. This can be in the simple application of a fully automatic system where recipes, groups and schedules are programmed into a dispenser and the system manages loading, mixing and distribution. Legacy equipment can interface into automated systems, retrofitting existing feeding systems to make them smarter and enabling automatic feeding every three to four hours.
Unmanned ground vehicles (UGVs) are a cutting-edge concept being pioneered by dairy equipment manufacturer Lely, who see robotic stable-hands as the innovative next step in smart farming. Andy Stewart, general manager at Lely Atlantic, says that traditionally a farmer could spend up to four hours a day feeding the herd, but with his company's Vector feeding robots a farmer can spend around three hours per week doing the same job.
"An automated crane fills the feeding robot with specific rations depending on the type of cow, for example if it is a younger group the robot will dispense different portions and compounds. A laser mounted to the front of the vehicle scans how much feed has been consumed so it knows when to redistribute. The robot is extremely flexible and can be programmed to distribute up to 16 varieties of rations." The UGV will add salt, protein, minerals and lime to a mix of dry, wet and fresh roughage several times a day to extend the life of the food it is distributing.
For industrial equipment manufacturers, the energy-saving secret of variable speed drives is a hot topic, and manufacturers of equipment for the smart farming industry are thinking along the same lines. The perception is that with increased automation comes increased costs, but there are methods which actively reduce energy consumption.
In the dairy farming process, the automated milking rotary transports the milk through the milking line with the help of a vacuum powered by a drive. In the past, the vacuum was required to be at a certain capacity to clean the milking system which is a much higher rate than that needed for general milking, at around 2-3 per cent of the pump capacity. A fixed speed drive meant the pump would be constantly running at full capacity, with air bled into the vacuum to reduce this when milking. To prevent energy wastage, DeLaval integrated a frequency-controlled vacuum pump with a variable speed drive, which has reduced energy consumption levels dramatically.
"Frequency control is something that has been used across many other industries, but it is the way that we use it that makes it unique to the dairy industry," says Manneke, "Less advanced models have to use shielded cables and special power supply, but our variable-speed drive can run on a regular power supply without electromagnetic disturbance." Heat recovery is another way to reduce energy consumption in smart farming, by reusing recovered heat from compressors used to cool the milk.
Future of smart farming
Some in the industry say that smart farming will move towards the segregation of production cows and recreational cows with a growing focus on health management, with the first type farmed predominantly indoors with extensive automation, and the other mainly farmed outdoors to produce a richer yield of milk. But with the burgeoning growth of world population, Stewart believes the real future of smart farming lies in skilled labour, with employees qualified to operate continuously evolving farm equipment and technology.
"One huge driver for smart farming is succession problems on farms; farmer's sons and daughters do not want to be hand milking and feeding cows. There is a long and prosperous future in smart farming, and to sustain it each nation must become more food dependant. Automated farming means improving product quality and the job that we do, and for many is the way forward." *