Why it matters
Without this system, irregular or hidden behaviours can go unnoticed, leading to stressed animals, welfare concerns, or failed accreditation inspections due to undetected welfare indicators or data gaps.
From design to safety
Design
AI supports exhibit and pool design
AI assesses swim patterns, social clustering, and night activity to inform pool layout, camera placements, lighting, and viewing angles for welfare and observer coverage. It models crowd flow, access routes for divers, and zones of attention for visitors, ensuring redundancy in camera coverage and fail-safes for night surveillance.
- Underwater camera array
- Infrared night-vision camera
- Hydrophone array
- Overhead pool camera
Build
Commissioning for multi-sensor feeds
During build, sensor calibration and data pipelines are validated, including camera alignment, microphone calibration, network bandwidth, synchronization with the data hub, and verification of time-stamped event streams. Field tests confirm cross-sensor correlation for trajectory and proximity metrics.
- Underwater camera array
- Infrared night-vision camera
- Hydrophone array
- Pool data hub server
Operation
Real-time monitoring and event scoring
Synthetic analyses read swim trajectories, social proximity, and nocturnal activity, generating welfare scores and event logs for on-shift staff review. Rulesets incorporate enrichment objectives, pair-bond recognition, and potential spectator impact during shows while preserving privacy.
- Underwater camera array
- Infrared night-vision camera
- Hydrophone array
- Overhead pool camera
Maintenance
Regular calibration and data integrity checks
Routine alignment checks, lens cleaning, sensor calibrations, data integrity audits, and firmware updates are scheduled. Data pipelines undergo integrity verification, clock skew checks, and redundancy testing for backup storage and failover.
- Underwater camera array
- Infrared night-vision camera
- Hydrophone array
- Overhead pool camera
Safety
Welfare and safety assurance workflows
Alarms and dashboards route welfare concerns to aquatic veterinarian and dive safety officer; logs support accreditation audits and corrective actions. The system enforces escalation to veterinary review, ensures dive team pauses if distress signals persist, and maintains regulatory-compliant documentation.
- Underwater camera array
- Infrared night-vision camera
- Hydrophone array
- Pool data hub server
Part by part
- 1
Underwater camera array
- Watch for
- Lens fogging or debris obstructing field of view during behavioral sessions.
- AI action
- Detects lighting, water clarity, and occlusions to trigger camera maintenance alerts.
- 2
Infrared night-vision camera
- Watch for
- Elevated noise in low-light, thermal bleed, or sensor saturation risks.
- AI action
- Alerts staff to low-light periods and adjusts exposure or activates IR.
- 3
Hydrophone array
- Watch for
- Drift in localization calibration causing inaccurate swim trajectory estimates over time.
- AI action
- Reads position data to flag deviations and triggers retraining of trackers.
- 4
Overhead pool camera
- Watch for
- Occlusion by multiple animals creating blind spots during social interactions.
- AI action
- Identifies proximity events and triggers enrichment or crowding alerts for staff intervention.
- 5
Infrared pool camera for night behavior
- Watch for
- Unusual rest patterns, pacing, or nocturnal vocalizations that may indicate distress.
- AI action
- Detects abnormal nocturnal activity and generates welfare alerts for veterinary review.
- 6
Pool data hub server
- Watch for
- Data latency or synchronization failures across cameras leading to incomplete events and misinterpretations.
- AI action
- Orchestrates multi-modal alerts and timestamps events for review by the care team.
Worked example
Catching a seal swim pattern anomaly before a show
Illustrative scenario with sample data.
An AI-driven monitoring system flags a deviation in a seal’s swim trajectory and social proximity during pre-show pool clearances, suggesting potential distress or fatigue.
Signal in
timestamp:2026-11-07T21:15:42Z; animal_id: SEAL-04; swim_speed: 1.8 m/s; proximity_to_others: 0.8 m; night_mode: true; camera_id: CAM-03; noise_level: 42 dB
System out
alert_id: ALERT-1287; welfare_flag: true; score: 72; recommended_action: veterinary review and enrichment adjustment; report_due: 2026-11-08 09:00.
Human sign-off: Aquatic veterinarian reviews welfare flag, then approves enrichment tweak and operational pause if needed.
How to roll it out
Phase 1
Weeks 1 to 4
Baseline data collection across all cameras with initial alert thresholds tuned.
Phase 2
Weeks 5 to 8
Pilot welfare dashboards and automated notes reviewed by staff during shifts.
Phase 3
Weeks 9 to 12
Full integration into routine operations with escalation to veterinary and safety officers.
Where people stay in charge
- Decisions about welfare risk levels must be signed off by the aquatic veterinarian and the dive safety officer.
- Only qualified staff may modify alert thresholds or approve changes to enrichment strategies.
- Critical safety alarms require immediate escalation and documented review by responsible personnel.
Questions owners and crews ask
What data feeds power the monitoring system in real time?
Real-time feeds come from underwater cameras, infrared cameras for night monitoring, a hydrophone array, and a central pool data hub server, all synchronized for coherent event analysis.
Who reviews AI alerts about welfare concerns?
Aquatic veterinarians review welfare alerts; the dive safety officer assesses safety implications, with the curator and senior aquarists confirming action plans.
How does the system handle nocturnal behaviour?
Infrared night cameras capture nocturnal activity; AI analyzes rest patterns and atypical movements, triggering welfare reviews when anomalies arise.
What records are produced for accreditation and audits?
Event logs, welfare flags, enrichment actions, and staff approvals are archived with timestamps for audits and regulator review.
Sources and further reading
- AMMPA: Standards and GuidelinesAccreditation standards for marine mammal pools, water quality, and training records used in audits.
- AMMPA: Aquatic Mammals JournalPeer-reviewed research on marine mammal behaviour, health, and husbandry, including nocturnal patterns.
- International Association for Aquatic Animal MedicineProfessional network and conferences for aquatic animal veterinary medicine and aquarium vets.
- WAZA: Animal-Visitor Interaction GuidelinesWelfare rules for animal-visitor interaction and how shows and spectators affect animals.
- The Marine Mammal Center: RehabilitationRehabilitation pool monitoring, medical care, and patient progress indicators for marine mammals in distress.
- HSE: Diving at WorkDiving at Work Regulations, approved codes of practice, and dive team safety duties for operators.
- GOV.UK: Keeping Zoo AnimalsZoo licence requirements, inspections, and record obligations for aquariums open to the public.
- BIAZA: AccreditationAccreditation of UK and Irish aquariums against welfare and conservation standards and inspection criteria.
- Aquatic Animal Life Support Operators (AALSO)Life support operator training, water quality practice, and certification for monitoring equipment upkeep.
- Species360: MissionAnimal records network with data standards for collection records and event data handling.
- EAZA: Animal WelfareWelfare assessment guidance and positive welfare practice for animals in zoos and aquariums.