calf movement trackers

A calf lounges at a South Dakota dairy. Movement trackers for calves have been shown to help spot sick calves earlier.

Identifying and treating disease in young calves as early as possible is crucial to maintaining the health of the calf, and also maintaining growth and performance.

New research out of Europe suggests that the use of accelerometers, or movement trackers, could help producers identify and treat sick calves nearly 10 days before the calf starts to show signs of illness.

“There is already evidence both in pigs and cows that you can predict the incidents of disease from behavioral changes several days before the farmer is able to identify these animals as a sick,” said Sergio Calsamiglia from the University of Barcelona, speaking at the International Precision Dairy Conference in Rochester, Minnesota.

Calsamiglia’s research focused on dairy bull calves, Friesian calves to be exact, but the research and information is transferable to calves in different production systems.

The research goal was to find ways to reduce antibiotic use by identifying and treating only the sick calves versus a mass treatment of a whole group and by identifying the disease early, requiring less treatment.

The study tracked 325 bull calves in four different groups spread out over the year. Each calf was given an accelerometer that monitored its movement, tracked its steps, lying time and visits to the feed bunk. It recorded how often the calf visited the bunk and how long it stayed at the bunk each time.

“Each group was monitored for nine weeks from arrival,” he said. “The first day, these animals were setup with a pedometer on the front right leg.”

Health checks were done daily. When a calf was deemed sick by the farmer, showing visible signs of illness, a vet was called to confirm the diagnosis. The day of diagnosis was considered day zero.

The researchers then reviewed the movements of that animal for the previous 10 days and watched the movements for the next 10 days.

“When an animal was sick, we identified what we called ‘three brothers.’ A brother was a calf that was in the same group, at the same days of life and was at about the same body weight at entrance to the farm,” he explained

The brothers’ movements were also reviewed for 10 days prior to and after day zero.

Throughout the whole study, 33 calves were diagnosed sick, about 10 percent. Of those 33 calves, 17 were diagnosed and treated for a respiratory disease. The other calves had non-specific symptoms and received broad spectrum antibiotics and anti-inflammatories.

“The final objective was trying to define by logistic regression if there is one day, one single day where I can say with the highest possible reliability that this animal is at risk of becoming sick, so I could try to develop an algorithm where the system would alarm you that these animals are at risk,” Calsamiglia said.

They used the movement information collected from the sick calves compared to the healthy calves to develop the program that would alert a farmer when a calf may be sick or at risk of becoming sick.

It was evident that even 10 days prior to being diagnosed, sick calves started showing signs in their movement. Sick calves visited the bunk 20 percent less than the healthy calves and spent less time at the bunk. The sick calves also took fewer steps throughout the day during the 10 days prior to diagnosis and the 10 days following treatment.

Trips to the bunk and time spent at the bunk recovered very quickly after treatment.

Once the program was developed that would track the calves’ movements and alert the producer to a sick calf, the movements of all 325 calves over the course of the study were reviewed by the program to test its accuracy.

The program had a false positive alert, saying a healthy calf was sick, 60 percent of the time.

“Which means that out of the 325 calves, we will have 70 calves with an alarm, but only 28 will be actually sick,” he said.

The other side of this is: how many animals that are actually sick, but the system does not alert for? Of the 255 calves that never alerted as sick, five were in fact sick at one point.

“With the system, we missed five animals out of the 33 [sick calves] and we will have a large number of animals, about 40 animals, that will not be actually sick,” he said.

When they asked the farmer that they were working with how he felt about this, he saw this as 250 calves he would not need to treat because he knew they were not sick. Versus treating all 325 calves because he couldn’t be sure.

The team is currently working on reviewing the data again and looking for ways to refine the system for more accuracy.

“The difference between healthy animals and a sick calve appeared at least 10 days prior and maybe we should’ve gone farther back in time,” Calsamiglia said. “Prediction may help in early preventive treatment system and it may be useful as an alarm system just to identify at risk animals.”

Sign up for our Weekly newsletter

* I understand and agree that registration on or use of this site constitutes agreement to its user agreement and privacy policy.