Why it matters
Without this system, breeding timing can drift, ovaries may be under- or over-stimulated, and foaling risk rises. Mis-timed coverings can disrupt cycles and reduce foaling success.
From design to safety
Design
AI-guided breeding design and layout
AI analyzes mare cohorts, teasing data, and facility layout to propose teases, ultrasound access points, and flow for breeding operations.
- teasing stall with automated teaser system
- transrectal ultrasound scanner
- follicle size measurement tool
- ovulation timing sensor
Build
Install data streams and verification
Structured data pipelines from teasing records, ultrasound devices, and calendar controls are installed and validated for accuracy and redundancy.
- teasing stall with automated teaser system
- transrectal ultrasound scanner
- follicle size measurement tool
- ovulation timing sensor
Operation
Real-time monitoring and decision support
Ongoing reads of teasing outcomes, follicle progress, and ovulation signals trigger breeding window suggestions and alert handlers.
- teasing stall with automated teaser system
- transrectal ultrasound scanner
- follicle size measurement tool
- ovulation timing sensor
Maintenance
Data integrity and device calibration
Regular calibration of ultrasound and sensors, audit of records, and updating breeding plans based on current cycle data.
- follicle size measurement tool
- ovulation timing sensor
- covering date calendar and AI planning module
- insemination kit and storage fridge
Safety
Biosecurity, welfare, and compliance checks
Sign-offs ensure mare welfare, licensing adherence, and isolation protocols during breeding interventions and data handling.
- covering date calendar and AI planning module
- insemination kit and storage fridge
- teasing stall with automated teaser system
- transrectal ultrasound scanner
Part by part
- 1
teasing stall with automated teaser system
- Watch for
- Wear or jam in teaser mechanism; inconsistent mare responses; biocontamination risk in area present.
- AI action
- Trigger safety pause and alert supervisor if teaser repeats without responses.
- 2
transrectal ultrasound scanner
- Watch for
- Probe wear, calibration drift, signal dropout during scanning; inconsistent follicle visualization.
- AI action
- Flag calibration due; prompt operator to redo scan and record image quality score.
- 3
follicle size measurement tool
- Watch for
- Unreliable size readings; image artifact affecting diameter estimates and trend over time.
- AI action
- Compute size trend; alert if rate exceeds threshold for impending ovulation window.
- 4
ovulation timing sensor
- Watch for
- Late or missed ovulation signal; sensor saturation; data gaps in timing.
- AI action
- Issue recommended breeding date window and confirm with attending vet.
- 5
covering date calendar and AI planning module
- Watch for
- Conflicting breeding windows across mares; data gaps in cycle history.
- AI action
- Produce consolidated breeding plan; lock approved dates for insemination teams.
- 6
insemination kit and storage fridge
- Watch for
- Temperature drift; shelf life warnings; semen inventory discrepancies.
- AI action
- Send restock and temperature drift alert; reconcile semen inventory.
Worked example
Catching a mare’s optimal breeding window before standing heat
Illustrative scenario with sample data.
A mid-sized stud yard with 28 mares and limited ultrasound access.
Signal in
maidenMare_id: M-204, teasingHistory: 3 events in 2 weeks, follicleSize: 9.2mm rising, ovulationSignal: pending, lastAIbreedingWindow: 14 days ago, calendar: B-Date: 2026-11-14, mare_birthparity: multiparous; last_tease: 2 days; cycle_history: 3 cycles.
System out
Breeding window 2026-11-22 to 2026-11-24; ultrasound check planned 2026-11-22; insemination 2026-11-23; post-ov check 2026-11-25; breeding order and team notes.
Human sign-off: Stud manager signs after vet review confirms follicle size and ovulation signal.
How to roll it out
Phase 1
Weeks 1 to 4
Baseline data capture from teasing, ultrasound, and records to calibrate AI model against current yard practices.
Phase 2
Weeks 5 to 8
Pilot AI-guided breeding windows with two priority mares; adjust thresholds for returning cycles.
Phase 3
Weeks 9 to 12
Full deployment across all mares, with continuous performance review and escalation paths.
Where people stay in charge
- Decisions on breeding window must be signed off by the attending equine vet with current approvals.
- All data handling must pass biosecurity review and keep mare privacy and traceability intact.
- If any sensor shows persistent drift, stop data collection until calibration is completed and revalidated.
Questions owners and crews ask
How does AI determine the best breeding window for a mare?
AI analyzes teasing responses, follicle growth, and ovulation timing to identify the most probable fertile window, then suggests covering dates and checks for the team.
What happens if the data is incomplete or a scan is poor quality?
The system flags gaps, requests repeat scans, and requires vet sign-off before finalising covering plans.
Can this integrate with our existing mare health records?
Yes, it links to patient records, vaccination, and exam history to align breeding timing with overall health status.
Who signs off on AI-generated breeding plans?
The attending equine vet, with confirmation by the yard manager and the licensing body if required.