This document outlines the technical requirements, sensor fusion philosophy, UX expectations, and strict engineering constraints for the Aevea hardware ecosystem. Our goal is to transition from "PetTech 2.0" (plastic GPS toys) to "PetTech 4.0"—a jewellery-grade, unified physical control loop focused on longevity intelligence and resting biometrics.
Current Contract Scope (Phase 1): Design and engineering for the Link Collar, Bond Leash Handle, and App v1.0 ONLY.
Note: Phase 2 actuation products (Nest, Stem, Flow, etc.) and deep AI learning loops are documented here purely for forward-compatibility context. Do not design Phase 2 products during this sprint, but ensure Phase 1 data architecture supports future integration.
Aevea is a distributed network where Sensing Nodes (Phase 1) capture resting metabolic output and spatial data, and Actuation Nodes (Phase 2) manage metabolic input and environment, all orchestrated by the continuously learning Command Center (App).
LINK COLLAR • BOND LEASH
MOBILE APP (IOS/ANDROID)
v1.0 manages hardware, spatial safety & live tracking. Future v2.0 manages exact caloric balancing and AI health predictions.
NEST BED • STEM FEEDER
The biometric spine of the ecosystem. Focus is placed on discreet intelligence, resting biometrics, and robust tracking.
Core Biometrics (SCG + Accel)
Primary reliance on Seismocardiography (chest micro-vibrations) to capture absolute Resting Respiratory Rate (RRR) and motion-derived cardiac inference, cleanly bypassing dense fur barriers.
Optical Evaluation
Evaluating PPG feasibility for RHR. If signal degrades heavily via fur, SCG acts as primary.
Thermal Integrity
Skin and Ambient temperature tracking with mandatory internal battery thermal auto-cutoff.
The physical connection point. Engineered for premium tactility and solid-state reliability, integrating directly with the Link collar via Tether Mode logic.
Tether Mode (Virtual Leash)
Utilizes ultra-wideband (UWB) to sync with the Link collar, establishing a secure, invisible perimeter for off-leash training in urban environments.
Active Recall Assist
Triggered via the handle trackpad. Sends distinct, trained haptic vibration patterns instantly to the Link collar to silently recall the pet.
Parasitic Link Charging
The Bond handle charges wirelessly via Qi. The leash line contains a conductive wire allowing the handle to parasitically charge the Link collar while physically tethered during walks.
Visual references detailing the intended hardware interface and spatial logic exclusively for the Tether Mode (Virtual Leash) feature.
The command centre for pet longevity. Turns biometric data into actionable care without relying on overly clinical or medical jargon in the consumer UI. The architecture must accommodate a freemium model, providing core safety and identity features for all hardware owners while reserving advanced longevity insights for intelligence subscribers.
Core safety feature. UI for setting spatial boundaries, haptic alerts, and managing Bond & Link provisioning.
Real-time location mapping pinged from the Link collar's integrated GPS module, activated upon emergency or breach.
Travel & Tenancy Ready. A verifiable, shareable digital record of identity, city permits, vaccinations, and health status for airlines and landlords.
Orthopedic Early Warning. AI analyzes stride length, cadence, and symmetry from the Link collar to detect early signs of pain or injury before they are visible.
Remote Clinical Dashboard. Owners can share a "Pro Link" granting their veterinarian temporary access to 30-day resting HRV, RRR, sleep, and gait trends to support clinical visits.
The app contains an AI continuously trained on accumulated data from all customers to generate proactive wellness warnings and health predictions.
The app tracks pet weight (via Nest scale) and calculates metabolic burn, making proactive suggestions for specific food dispense amounts to the hardware (via Stem/Flow logic).
While the user-facing application maintains a clean, minimal, Apple-like simplicity (avoiding medical phrasing), the underlying app architecture must support deep data utility. The database layer must be constructed to facilitate: Longitudinal health modeling, Cross-cohort data aggregation, potential Federated Learning pipelines, and an API layer for future enterprise pharmaceutical licensing.
OUT OF SCOPE: The following hardware is provided strictly for contextual understanding of the overall ecosystem logic. No design deliverables are required for these items in Phase 1.
AI Feeder
AI precision feeder for exact caloric input tracking, multi-pet recognition, and appetite fluctuation alerts.
Smart Fountain
Smart hydration fountain capturing exact water intake to identify early signs of kidney or urinary issues.
Biometric Bed
Biometric bed featuring an integrated scale for passive daily weight tracking, sleep analysis, and thermoregulation.
Air Monitor
Environmental monitor correlating indoor air quality (AQI) and allergens with respiratory biometrics.
Smart Bowls
Smart bowls for passive calorie and hydration logging.
Recovery Bed
Enclosed recovery bed designed for post-surgery healing and anxiety calming.
Smart Harness
Clinical smart harness capturing ergonomic biometrics, posture tracking, and kinetic output metrics.
Auto Litter
Auto cat litter robot with AI camera to detect pee/poop and smart colour-changing pH litter.
Aevea is positioned strictly as a consumer wellness and safety device in Phase 1.
Rule 1: No diagnostic medical claims are to be made in hardware or app UX.
Rule 2: However, the underlying data infrastructure must be engineered to rigorous standards to allow for future clinical validation studies and FDA/RCVS discussions.