A state-of-the-art robotic dairy unit at Harper Adams University gives a glimpse at what the digital technology and infrastructure units of the future may feature.
Harper Adams University has a long-standing reputation for research in agriculture, with the latest projects focusing on the future of dairy farming.
Teaming up with the Agri-EPI Centre, which focuses on engineering, precision and innovation, the project also draws upon industry manufacturers and service providers involved in building infrastructure and software to enhance the performance and welfare within the dairy industry.
Professor Mark Rutter heads the department at the university and explains the thought process behind the project.
He says: “We now have some 80 companies involved with the project, which is looking into all aspects of the dairy industry.”
He adds the end goal is to homogenise the software used in the industry and hopefully encourage all manufacturers to develop ‘plug and play’ systems, which can be implemented on any farm.
“If you look at the tractor industry for example, IsoBus has revolutionised the way manufacturers develop their products and in theory you can select the best implements from any manufacturer, knowing they will work with your tractor,” he says.
“However, in the dairy industry, you can count on one hand the manufacturers that offer a complete robotic solution, so we are trying to create a platform where any manufacturer’s machines will work with another, with minimal modification.”
At the heart of the project is a 50-cow shed measuring 60 by 24 metres (197 by 79ft). Housed yearround, milked through a robotic parlour and bedded on cubicles with sawdust, every aspect of the building has been meticulously designed.
“The building was designed to be flexible,” Prof Rutter says. “It is bigger than is needed for the number of cows, but areas are shut off to replicate real world conditions.
“The plan is to house all the cattle needed for this unit under one roof, from reared calves, to bulling heifers through to dry and lactating cows.”
The equipment in the shed has also been meticulously thought out, including the plastic cubicles.
“We know that given the choice, cows would rather be lying on grass, but we do not know why,” he adds.
“Is it the softness of the grass, is it the amount of room they can have to stretch out, or is the wide field of view they get?
“To find out, the plastic cubicles mimic the amount of room a cow would use outside, and mattresses topped with sawdust hopefully give them the softness grass would provide.
“The shed was designed so that wherever a cow is in the shed, it can have an unobstructed view to the outside.”
TAKING centre stage in the unit is the DeLaval VMS robotic parlour.
DeLaval already has a sophisticated data recording suite and was open to the idea of gathering more data to enhance production and welfare for farmers.
Called Herd Navigator, the system uses neck collars with accelerometers incorporated to relay information to the system.
A complex algorithm fed from the neck collar data, as well as management data, decides if milk should be tested for various reasons.
Analysis of the milk for lactose dehydrogenase, progesteroneand beta hydroxybutyrate, can indicate the start of oestrus, mastitis and ketosis, giving prior warning before visual signs become apparent.
The robotic parlour will also predict the yield of the cow and individual quarters through another algorithm, notifying the manager if target is not met, the cow has not been milked in a period or if it persistently kicks units off.
Herd Navigator will also put a marker on the cow if it detects anything untoward in the milk, so a management decision can be made. On the exit of the parlour is a 3D camera which scans the cow and automatically gives it a condition score.
Although at present the data is not directly relayed to the in-parlour feeder, it is sent to the manager who can decide to increase or decrease the amount of feed administered at milking, aiming to create a level batch of cows, making it easier to manage and feed them with the partially mixed ration the cows are fed.
TO help farmers achieve better and more consistent welfare and profitability, the university, Muller, Kingshay and Ice Robotics have embarked on a project called Dairy Animal Sensor Integrated Engineering.
Prof Rutter says: “The project will explore integration with existing farm systems and equipment such as milk meters and feed dispensers, maximising the systems farmers already have.
“This will allow farmers to take swift action to address health conditions such as fertility problems, lameness and mastitis before the situation becomes more serious and costlier.”
Research into behavioural profiles for other diseases and conditions is also being undertaken.
Mastitis and ketosis are two of the biggest problems dairy farmers face, with treatment both costly and time consuming.
The research aims to pick up on any behaviour which deviates from the individual animal’s norm.
This is a new way of looking at animal behaviour, which is usually based on deviations from the group’s average, and not on a per cow basis.
It is thought that there are signatures that become apparent in each animal’s behaviour, such as restlessness, sedation and changes in gait which could indicate a particular condition, before it is showing clinical signs.
It is hoped that all the software within the building will communicate with each other, helping to get quicker diagnostics.
For example, with mastitis, if the IceQube detected a change in the animal’s behaviour, that it knew had the mastitis infection signature, it would instruct the robotic parlour to carry out a milk test for high cell counts, and the cow segregated for treatment if it were the case.
If ketosis were an issue, extra feed or a more energy dense feed could be administered at the parlour.
POTTERING about the shed is a DeLaval RS450S robotic scraper.
The scraper has a pre-planned route, with location tags to let the machine know where it is.
It will typically work for 20 hours per day, with the rest of the time spent charging and refilling the onboard water system.
Apart from occasionally getting stuck, Prof Rutter says the machine has been reliable.
It is the first scraper used in the shed, but the design of the alleys and cubicles was done to facilitate the use of robots.
In the future, Prof Rutter says infra-red technology could be used to identify the hot spots where dung has been freshly dropped, reducing the time spent travelling where no manure is.
Similar in principal is the unit’s Joz Moov-Pro feed pusher, which too uses transponders in the floor and pre-planned routes, as well as 3D mapping to navigate the feed passages. Increased feed intake by the whole herd, but particularly lower ranking cows, and consequently high milk yields is the biggest advantage of robotic feed pushing.
The ability to apply liquid concentrates such as molasses to further enhance intake could be an option in the future.
THERE are hundreds of pedometer and accelerometer ankle sensors on the market, but Ice Robotics’ IceQube was chosen to be used in the unit.
Now in its third generation, the sensor records activity including standing and lying, as well as oestrus detection.
One of the reasons the IceQube was chosen is because the company is exploring algorithms and sensors which detect a change in a cow’s gait, before it becomes apparent to the stockman, catching lameness in its very early stages.
So far, the team thinks it may have found a pattern which identifies cows positive for Johne’s disease.
Prof Rutter explains that data has initially shown JD5 cows spending up to two hours more standing, compared to JD0 cows.
The hypothesis here is that the Johne’s-positive cows are spending more time eating.
This would coincide with peak lactation, with the cow’s digestion prioritising milk production over maintenance, possibly explaining Johne’s-positive cows becoming emaciated, even though their appetite is normal.