Farm sensor data: connect devices and records. Build a consistent record of device identity, measurement time, measurement type and unit, linked animal and review data frequency, missing readings, unusual readings over time.
Good recordkeeping is more than collecting data. It connects each event to the right identity, date and source so you can review what happened. This guide offers a practical starting point for herd managers and field teams. It supports record organization; it does not replace expert assessment, official records or professional decisions.
1. Define the scope and responsibility
Identify the business, parcel, animal, flock, hive or patient the record belongs to. Include the event date and responsible person. Consistent names and identifiers prevent several people from recording the same event under different identities.
Start with a small sample from your existing records. Try one week and check that everyone uses fields with the same meaning. Keep identity, date, quantity and unit complete without making unused fields mandatory.
2. Include the essential fields
Start with device identity, measurement time, measurement type and unit, linked animal. Use short, clear labels. Always store the unit alongside a quantity: a number alone cannot distinguish litres, kilograms, items, area or currency.
The event date may differ from the entry date. Preserve the actual event date when recording yesterday’s work today. When correcting a record, retain who changed it, when and why.
| Record field | Why it matters | Check |
|---|---|---|
| device identity | Matches the correct record | Is device identity clear and verifiable? |
| measurement time | Clarifies time and scope | Is measurement time clear and verifiable? |
| measurement type and unit | Enables comparison | Is measurement type and unit clear and verifiable? |
| linked animal | Retains the source | Is linked animal clear and verifiable? |
3. Review daily records each week
Daily entry and weekly review serve different purposes, even if one person does both. Enter promptly and accurately, then review missing dates, unusual changes and open tasks. A missing record is not the same as a measured zero.
Ask whether records are complete, whether there is a significant change from the previous period and whether anything awaits assessment by a responsible person or specialist. Connect the record to the work plan.
4. Follow meaningful indicators
Begin with data frequency, missing readings, unusual readings. Use the same definitions and units each period. When animal count, parcel area, flock age or recording duration changes, compare rates or per-unit values alongside raw totals.
An increase or decrease does not explain its cause. First check completeness, date ranges and units; then consider environmental, equipment, market or health context with the appropriate specialist.
5. Avoid common recording mistakes
Each integration requires the device brand, model, data access and field testing. For urgent, uncertain or regulated matters, contact the appropriate professional without waiting to finish entering a record.
- Do not create the same identity twice with different spelling.
- Store kilograms, litres, count, area or currency alongside quantities.
- Keep planned operations separate from completed events.
- Distinguish missing values from measured zero values.
- Link photos and documents to the correct record and date.
- Use individual user accounts rather than shared passwords.
- Retain the reason for corrections and the responsible person.
6. Start with a practical recordkeeping check
Create one sample record in the relevant available module and check that you can find it again using device identity. Add a second dated event, then check the chronology, notes, documents and responsible person. Ask another authorized user to locate the same record.
After a week, review data frequency and missing readings. If the results differ from your trusted source, check identities, dates and units before adding more records. Apply the same method to other records once it is consistent. See the feature catalogue for modules currently available in Tarliva.
Reliable records start with the correct identity, an explicit date, a visible unit and a responsible person.
