SMART FARMING ▶▶▶
Deep learning for efficient dairy farming
BY REBECCA KWAKMAN T
erry Canning, co-founder and CEO of CattleEye, un- derstands these challenges like no other. Canning tells Dairy Global more about his company and its mission to create the world’s first autonomous
livestock monitoring platform.
Deep learning A repeat entrepreneur, Canning grew up on a dairy farm in County Armagh (Northern Ireland), studied engineering and spent nine years working in telecommunications and cloud computing. In 2004, Canning applied his knowledge of cloud computing to the livestock industry and founded Farm- Wizard, which grew to manage about 4 million cattle. After selling FarmWizard to the Wheatsheaf group in 2015 and exit- ing the company completely in March of 2019, Canning de- cided to take a new approach to livestock monitoring: “My previous business allowed people to record information about livestock by using apps, but if you wanted to record movement or record the animal, it was always reported by humans,” Canning explains. “What I want to do with CattleEye is provide a way to do this automatically so that the farmer doesn’t have to be involved.” CattleEye uses video analytics powered by deep learning artificial intelligence (AI) to deliver insights to farmers.
Latest developments CattleEye’s technology makes use of the latest developments in the industry. “The computing power and the type of tools that enable machine vision to work correctly have now be- come available. Tools, such as Amazon Webservice, and better cloud computing capability allow us to develop AI that can make use of video analytics and derive specific insights from those,” Canning continues. CattleEye took off when Canning came across co-founder Adam Askew, who had spent 12 years working with the same kind of technology in cancer detection. Once Canning and Askew had founded CattleEye, they raised about € 1 million in investments and were joined by Askew’s team, made up of six people who were working on this technology, which had just been acquired by Philips. “Once we were up and running, we started with a very small farm of just 90 cows and used lots of cameras, which were
As the world population grows and the demand for livestock protein increases, farmers are trying to find new ways to up their production. Sustainable dairy farming, however, is not just a matter of farming more animals to meet the rising demand. With legislation limiting emissions from livestock, the future of dairy farming lies in increased efficiency rather than expansion.
‘learning’ how to work. Now, we’re monitoring 11,000 cows,” Canning adds. “While we were building up to that we did some work with the University of Liverpool, we brought on another farm in Dorset, and we just brought on our first farm in the US last month,” Canning says, and goes on to explain how CattleEye’s technology uses AI where farmers would previously have used their own eyes: “We are recording lots of videos and put- ting in expert scores – done by vets who go to the farms to score the cows – which we then use to train our neural net- works. These networks can look at a video of a cow, extract data, identify it and offer insights on factors such as mobility scoring and body condition scoring. “If we take mobility scoring, for example, the AI can use the video to identify lame cows early on. The farmer can put in a threshold score between 0 and 100, with 0 being the best and 100 being the worst. If a cow scores over the threshold,
Terry Canning, co-founder and CEO of CattleEye.
▶DAIRY GLOBAL | Volume 7, No. 4, 2020
25
PHOTO: CATTLEEYE
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28
