USING VISION RECOGNITION TO IDENTIFY SPECIFIC PROBLEMS AND CAPTURE INSIGHTS IN ANIMALS w/ZOETIS

APPROACH

View fullsize

In a partnership with Zoetis and Google X, and using Google Glass, Farmwave was asked to leverage our AI and vision compute capabilities to visually inspect and provide visual feedback on animals in their environment.

OBJECTIVE

Build multiple capabilities for gathering visual feedback in sows and their offspring within their own environment. The tools included:

• visual classifiers for anomalies related to illness and disease

• scanning of ear-tags to automate pulling health records

• weight estimation through a silhouette pattern recognition given the visual spect ratio of Glass at 2.4 meters wide and 6 meters forward

• use voice recognition to record and capture data completely hands-free

View fullsize

RESULTS

The movement of animals, specifically that of pigs, made it difficult to capture quality broad-side imagery for weight measurement. Larger, slower, animals would have been a better for this particular pilot. Physical defects and ear-tag scanning was pretty elementary and exceeded expectations. Voice recognition proved to be valuable for farmer input and data capture, however the lack of wireless connectivity and access to broadband connection limited both the recording of data beyond local Glass storage, as well as putting records from ear-tag scans. When connectivity was present, the system exceeded expectations.

KEY LEARNINGS

There is a lot of room to grow when it comes to the animal health sector of agriculture. From daily farmer tasks to veterinary handling, Farmwave has proven the value behind hands-free age recognition and data insights using vision recognition.