The problem
Dairy animals often show illness through subtle changes in activity and rumination days before visible symptoms. Early detection protects both animals and income.
What to build
Develop a collar-worn sensor pipeline that learns each animal's normal behaviour and raises an alert when activity deviates in a way associated with illness or heat.
Success looks like
Alerts that precede farmer-noticed symptoms in a field trial, a false-alarm rate small enough to be trusted, and a herd dashboard.
Suggested technologies
- Accelerometer data
- Anomaly detection
- LoRa
- Edge inference
- Time-series analytics
An original, independent statement. It is not an official Smart India Hackathon problem statement, and Arventiq Labs is not affiliated with or endorsed by that programme. The scenario is illustrative and does not describe any specific organisation. Free to use with credit.