The bionic farmer
Tuesday, 1 August 2017
VetScript Editor's pick - August
More than ever, dairy farmers are turning to technology to aid their everyday animal management. But with some of the latest gadgets able to signal the onset of an infection before an animal even shows clinical signs, where does that leave the veterinarian? Mirjam Guesgen investigates.
Daisy stands chewing her cud in the milking shed. Little does she know that she is being closely watched. In an instant, an infrared camera scans her udder to reveal she is two degrees warmer than yesterday. While she shows no other clinical signs yet, she is actually in the early stages of a mastitis infection.
Fifty kilometres away, the local veterinarian’s cellphone buzzes as she receives Daisy’s scan via email. In a few hours, the cow will be tended to first during the veterinarian’s daily farm visit.
New Zealand dairy farmers are under pressure to ensure the health of their cows. The lowest recorded milksolids payout in a decade (DairyNZ, 2016) and growing consumer interest in animal welfare (Barkema et al., 2015) mean farmers cannot afford to have sick animals in their paddocks.
But with herds larger than ever, and the cost of employing labour rising (DairyNZ, 2016), it’s a recipe for health issues going unnoticed.
In response, New Zealand is helping to develop and is implementing technology to monitor cow health and wellbeing. The purpose of the technology is to automate and enhance traditional farming practices. This allows farmers to see aspects of their animals’ health that the human eye simply can’t pick up, or that an inexperienced or overworked eye may overlook. This ability signals a looming change in the role of the herd veterinarian.
Farmers have available to them a cornucopia of health information for each cow, data from scanners that can cover nearly every centimetre of an animal’s body. On the low-tech end of the spectrum there are leg pedometers, rumination collars and ear tags, all designed to monitor activity, with some giving health alerts, such as when a cow isn’t eating enough.
A high-tech example is Swedish livestock technology company DeLaval’s automated body condition scanner. A 3D camera mounted above the sorting gates takes density and height measurements of a cow’s hindquarters and determines her body condition according to her breed. The information is available to the farmer, and is then used to decide best management practices, such as moving poor-condition cows to a fresh break.
Another technique, called kinematics, is straight out of The Lord of the Rings. Get a cow to walk past a series of 3D cameras or over a force plate and you can obtain precise characteristics of her gait, such as stride length and rate, limb rotation and back arch. Run those parameters through a computer algorithm and the system will automatically tell you when a cow is lame with 96% accuracy (Poursaberi et al., 2010). It’s even possible to predict which animals will later become lame by detecting the slightest variations in how they move (Viazzi et al., 2013).
Another technology, infrared thermography (IRT), uses thermal cameras mounted in a rotary shed or on a drinking station to measure temperature variations in the eye, hoof and udder. If there’s dysfunction in the thermoregulatory systems inside, as when a fever is triggered to fight infection, it will show through these thermal windows.
IRT has been able to detect digital dermatitis, FMD, bovine respiratory disease and mastitis infections (Thermography, 2013) up to six days before an animal shows any clinical signs.
IRT is also more accurate in predicting which animals will develop illnesses a week later, compared with just checking for visual signs (Schaefer et al., 2007). The advantage is that infected animals can be quickly isolated and treated to prevent disease spread and minimise suffering.
Unlike the body condition scanner, kinematics and IRT are still in the development phases. So are these technologies more science fiction than farming tool?
Al Schaefer, who is commercialising the IRT system through his Canadian company Animal Inframetrics, says the technology could be available as early as this year. “The two big issues for us are cutting costs and making it more user friendly. In the [past year we’ve seen] solutions to both of those, which is very encouraging.” However, he adds, they desperately need more engineers and computer scientists to meet the interest expressed.
Al has worked closely with Kiwi animal welfare scientists from AgResearch and DairyNZ since the early 1990s. Together they’ve championed many of the proof-of-concept studies for IRT’s use in disease detection in both New Zealand and Canadian cows. More recently he collaborated with Hamilton based company InterAg to trial the IRT system as a way to detect oestrus in dairy herds in Waikato.
New Zealand scientist Mairi Stewart researched IRT extensively with Al, through her time both at AgResearch and with InterAg, but is more sceptical than he is when it comes to commercialising the technology.
“It’s a lot harder to go from the proof of- concept stage to the commercial stage than people think. It’s all well when you’ve got lovely clean cows set up in the positions you want, but try putting them out in the muck in the middle of winter in New Zealand and you’ve got a whole other situation. There are ways of dealing with that in the data, but it’s a lot of work.”
These emerging livestock technologies come from centuries of farming practice, where well-trained personnel and a close relationship between person and animal were the diagnostic tools of choice. Seeing a cow fall behind the herd on her way to milking might have indicated lameness, and palpation and visual assessment of a few members of the herd could have given you an idea of body condition score overall. Times are changing, not necessarily because observing an animal is becoming an obsolete method of catching problems, but because financial pressures mean farmers need to do this more efficiently and accurately.
Peter Kemp, who has been in the agriculture sector for more than 30 years, says the visual model worked when herd sizes were small and farmers could afford to hire skilled labour to conduct checks.
“One of the challenges for farming is that the costs keep creeping up, so it’s all about efficiency,” says the Massey University professor. “There’s pressure to minimise labour and, in systems like that, if someone isn’t really experienced they miss things.”
It’s all about detecting problems before they arise, accurately, on an individual basis and with minimal time or cost. The technology is not to replace people – it’s to enhance them.
But Mairi warns of potential ‘data overload’, where the information these systems give is either too complicated to understand and make meaningful decisions from, or too vast. This could see veterinarians taking on more of an analyst or teaching role. Dairy cattle veterinarian resource manager for the NZVA Neil MacPherson says there is an opportunity for veterinarians to become advisors who can sift through the avalanche of data, prioritise it and make recommendations.
He sees farmers doing their own diagnosis and treatment using the new technology alongside annual treatment plans prepared by veterinarians.
Peter sees the future veterinarian going remote. They will receive snapshots of alerts generated by the technology on their smartphones or laptops, along with a note or photograph from the farmer. A remote approach could increase efficiency, because veterinarians will know exactly which cows to examine on their site checks. A similar tactic is already being employed in human health through telemedicine, aided by applications such as WhatsApp.
Al Schaefer agrees. “What this [technology] is, of course, is simply a tool. It’s a tool that will make veterinarians’ and farmers’ lives easier and more profitable.”
The key, however, is that no technology is completely ‘hands off’. It all still requires human judgement – be that farmer or veterinarian – to be effective.
“Like any data, you have better information to make your diagnosis,” says Al.
Barkema H, von Keyserlingk M, Kastelic J, Lam T, Luby C, Roy J, LeBlanc S, Keefe G, Kelton D. Invited review: changes in the dairy industry affecting dairy cattle health and welfare. Journal of Dairy Science 98, 7426–45, 2015
New Zealand Dairy Statistics. Accessed on 12 June 2017 from www.dairynz.co.nz/media/5416078/nzdairy- statistics-2015-16.pdf
Jabbar KA, Hansen MF, Smith ML, Smith LN. Early and non-intrusive lameness detection in dairy cows using three-dimensional video. Biosystems Engineering 153, 63–9, 2017
Poursaberi A, Bahr C, Pluk A, Van Nuffel A, Berckmans D. Real-time automatic lameness detection based on back posture extraction in dairy cattle: shape analysis of cow with image processing techniques. Computers and Electronics in Agriculture 74, 110–19, 2010
Schaefer AL, Cook NJ, Church JS, Basarab J, Perry B, Miller C, Tong AK. The use of infrared thermography as an early indicator of bovine respiratory disease complex in calves. Research in Veterinary Science 83, 376–84, 2007
Thermography. Current status and advances in livestock animals and in veterinary medicine. Edited by F Luzi, M Mitchell, LN Costa and V Redaelli. Brescia Foundation, Italy, 2013
Viazzi S, Bahr C, Schlageter-Tello A, Van Hertem T, Romanini CE, Pluk A, Halachmi I, Lokhorst C, Berckmans D. Analysis of individual classification of lameness using automatic measurement of back posture in dairy cattle. Journal of Dairy Science 96, 257–66, 2013