WALK over weighing technology is accurately predicting calving dates to 92 per cent within the week and 84pc within three days.
Accurate estimations of date of birth are a vital component for cattle breeders looking to evaluate growth rates and maternal fertility parameters such as days to calving.
The results in northern research herds are being replicated in commercial settings using the walk over scales during the calving season.
The calving date detection achievements are part of a five-year research project at CQUniversity, co-funded by the Meat & Livestock Australia Donor Company, aimed at increasing the uptake of performance-recorded genetics in extensive beef enterprises through the use of automated livestock management systems to measure key traits.
Researchers say the future for genetic improvement in northern Australian beef herds requires more cattle to have more accurate, more frequent and more reliable performance measures.
CQU researcher Lauren O'Connor said automated capture of calving dates had been a key focus of the project's DataMuster tool, which links hardware and software to deliver fast, simple and accurate monitoring of cattle performance in remote locations.
Now in its second year, DataMuster is currently managing walk over weighing information from 27 stations, close to 7000 cattle and almost half a million individual weight measurements.
Dr O'Connor said the aim of the DataMuster system was to produce meaningful information that could be used to make quick and informed decisions.
An algorithm had been developed that uses a cow's average weekly weights throughout the calving season to identify the most likely week of calving. It then searches the daily weight data to find the date of calving.
A trial at the Belmont Research Station near Rockhampton involving 112 cows showed the algorithm accurately predicted 92pc within the correct week of calving, 73pc within one day and 84pc within three days.
"This year we had a crack at commercial application with 291 cows. While we encountered many problems and were only able to obtain a calving date for 35 cows, it did yield similar results to those obtained in our research station trial," Dr O'Connor said.
"The algorithm displays good potential but there are still challenges around how to optimise cattle use of walk over weighing.
"The key message in regards to determining calving dates using walk over weighing is it takes a team with expertise in the technology, cattle handling and data science.
"It's not a case of just sticking a walk over weighing unit in the paddock and expecting good results."