Chapter 03: Data Collection APIs

WIA-SENIOR-007 Senior Wearable Standard | Data Architecture

The Data Pipeline

Raw sensor data is useless without a robust collection, transmission, and storage infrastructure. The WIA-SENIOR-007 Data Collection API defines standardized interfaces for streaming health data from wearable devices to cloud platforms, mobile applications, and healthcare systems. This chapter explores the complete data pipeline from sensor sampling to secure cloud storage, covering real-time streaming protocols, offline buffering strategies, data compression techniques, and HIPAA-compliant security measures.

For senior health monitoring, data collection must balance conflicting requirements: real-time responsiveness for emergency detection, power efficiency for multi-day battery life, reliability despite intermittent connectivity, and privacy protection for sensitive health information. The standard achieves this through a layered architecture with intelligent data prioritization and adaptive transmission strategies.

API Architecture Overview

The WIA-SENIOR-007 data collection architecture consists of four primary layers, each with specific responsibilities and performance characteristics:

Layer 1: Device-Local Data Collection

The embedded firmware continuously samples sensors at configured rates (typically 25-125 Hz for PPG, 50-100 Hz for IMU), performs on-device preprocessing, and buffers data in local flash memory. This layer handles power management through intelligent duty cycling and provides the foundation for real-time emergency detection that operates even when disconnected from the cloud.

Layer 2: Device-to-Gateway Communication

Bluetooth Low Energy 5.0 or Wi-Fi 6 provides wireless connectivity to a smartphone gateway or home hub. The protocol uses efficient binary serialization, delta compression, and priority queuing to minimize power consumption while ensuring critical alerts transmit immediately. Automatic reconnection and conflict resolution handle network disruptions gracefully.

Layer 3: Gateway-to-Cloud Synchronization

HTTPS/TLS 1.3 connections secured with mutual authentication and certificate pinning transmit data from gateways to cloud infrastructure. The layer implements exponential backoff for retries, intelligent batching to reduce connection overhead, and differential sync to avoid redundant transmissions. Rate limiting and quotas prevent accidental data charges from harming users with limited mobile plans.

Layer 4: Cloud Data Processing & Storage

Server-side infrastructure receives data streams, validates integrity, performs advanced analytics, stores time-series data in HIPAA-compliant databases, and distributes insights to authorized healthcare providers, family caregivers, and third-party applications through secure APIs. This layer handles scale, redundancy, and compliance requirements.

API Layer	Protocol	Latency Target	Data Rate	Power Impact
Device Sampling	Embedded firmware	<10ms	10-50 KB/minute	Baseline (sensor power)
BLE Communication	BLE 5.0 GATT	<100ms	1-5 KB/s sustained	+15-25% battery drain
Wi-Fi Communication	Wi-Fi 6 (802.11ax)	<50ms	10-100 KB/s burst	+40-60% during transmission
Cloud Sync (HTTPS)	RESTful HTTPS/2	200-500ms	Variable (batched)	Gateway-dependent
Emergency Alert	WebSocket (persistent)	<200ms end-to-end	Small bursts (<1 KB)	Minimal (rare events)

RESTful API Specification

The WIA-SENIOR-007 cloud API follows REST principles with JSON payloads, versioned endpoints, and comprehensive error handling. All endpoints require OAuth 2.0 authentication with device-specific or user-specific tokens.

Core API Endpoints

// Base URL: https://api.wia-senior-007.org/v1

// ============================================
// Health Data Submission
// ============================================

POST /devices/{deviceId}/health-data
Content-Type: application/json
Authorization: Bearer {device_token}

Request Body:
{
  "timestamp": "2025-01-15T14:32:45.123Z",
  "sequence": 12847,
  "data": {
    "ppg": {
      "heartRate": 72,
      "hrv": {
        "rmssd": 42.3,
        "sdnn": 58.1
      },
      "spO2": 97,
      "confidence": 0.94,
      "irregularRhythm": false
    },
    "imu": {
      "stepCount": 4521,
      "activityLevel": "LIGHT_ACTIVE",
      "postureQuality": 0.87
    },
    "temperature": {
      "skin": 33.2,
      "ambient": 22.5
    }
  },
  "alerts": [
    {
      "type": "HEART_RATE_HIGH",
      "severity": "MEDIUM",
      "timestamp": "2025-01-15T14:30:12.000Z",
      "value": 115,
      "threshold": 100
    }
  ]
}

Response (200 OK):
{
  "status": "accepted",
  "dataId": "hd_a7f3k2m9p1q4",
  "processed": true,
  "warnings": []
}

// ============================================
// Retrieve Historical Data
// ============================================

GET /users/{userId}/health-data?start={ISO8601}&end={ISO8601}&metrics=heartRate,spO2
Authorization: Bearer {user_token}

Response (200 OK):
{
  "userId": "usr_k3j2h1g9",
  "deviceId": "dev_x7y8z9a1",
  "timeRange": {
    "start": "2025-01-15T00:00:00Z",
    "end": "2025-01-15T23:59:59Z"
  },
  "data": [
    {
      "timestamp": "2025-01-15T00:00:00Z",
      "heartRate": 58,
      "spO2": 96
    },
    {
      "timestamp": "2025-01-15T00:01:00Z",
      "heartRate": 59,
      "spO2": 96
    }
    // ... more data points
  ],
  "statistics": {
    "heartRate": {
      "min": 52,
      "max": 115,
      "average": 68.4,
      "stddev": 12.3
    },
    "spO2": {
      "min": 94,
      "max": 99,
      "average": 96.8,
      "stddev": 1.2
    }
  }
}

// ============================================
// Emergency Alert Webhook
// ============================================

POST /webhooks/emergency-alert
Content-Type: application/json
Authorization: Bearer {service_token}

// Server sends this to registered emergency contacts
Payload:
{
  "alertId": "alert_p9q8r7s6",
  "userId": "usr_k3j2h1g9",
  "deviceId": "dev_x7y8z9a1",
  "type": "FALL_DETECTED",
  "severity": "CRITICAL",
  "timestamp": "2025-01-15T14:45:32.123Z",
  "location": {
    "latitude": 37.7749,
    "longitude": -122.4194,
    "accuracy": 15, // meters
    "address": "123 Main St, San Francisco, CA"
  },
  "context": {
    "impactMagnitude": 3.2, // g-force
    "immobilityDuration": 45, // seconds
    "userResponseAttempts": 2,
    "userResponded": false
  },
  "recommendations": [
    "Call emergency services immediately",
    "Check on user in person if nearby",
    "Review user's medical history for relevant conditions"
  ]
}

Data Streaming for Real-Time Monitoring

For applications requiring continuous monitoring (e.g., hospital telemetry, clinical trials), WIA-SENIOR-007 supports WebSocket streaming that pushes data to clients as soon as it arrives at the cloud platform. This enables real-time dashboards and immediate alert notifications.

// WebSocket Streaming API
// Connect to: wss://stream.wia-senior-007.org/v1/devices/{deviceId}/stream

// Authentication via query parameter
wss://stream.wia-senior-007.org/v1/devices/dev_x7y8z9a1/stream?token={stream_token}

// Messages sent from server to client:
{
  "type": "HEALTH_DATA",
  "timestamp": "2025-01-15T14:32:45.123Z",
  "data": {
    "heartRate": 72,
    "spO2": 97,
    "activityLevel": "RESTING"
  }
}

{
  "type": "ALERT",
  "alertId": "alert_p9q8r7s6",
  "severity": "HIGH",
  "alertType": "IRREGULAR_RHYTHM",
  "message": "Possible atrial fibrillation detected",
  "timestamp": "2025-01-15T14:33:12.456Z"
}

{
  "type": "DEVICE_STATUS",
  "batteryLevel": 42,
  "signalStrength": -67, // dBm
  "firmwareVersion": "2.3.1",
  "lastSync": "2025-01-15T14:32:50.000Z"
}

// Client can send control messages:
{
  "type": "SUBSCRIBE",
  "metrics": ["heartRate", "spO2", "alerts"]
}

{
  "type": "UNSUBSCRIBE",
  "metrics": ["activityLevel"]
}

Data Security & Privacy

Health data is among the most sensitive personal information, and seniors are particularly vulnerable to privacy violations. WIA-SENIOR-007 mandates comprehensive security measures at every layer of the data pipeline.

Encryption Requirements

• Data in Transit: TLS 1.3 with certificate pinning for all cloud communications. BLE connections use AES-128 CCM encryption with per-session keys rotated every 24 hours.
• Data at Rest: AES-256-GCM encryption for all stored health data, with keys managed through hardware security modules (HSM) or cloud key management services (KMS) that meet FIPS 140-2 Level 3 standards.
• Data in Processing: Encrypted memory regions for sensitive computations. Differential privacy techniques for aggregate analytics ensure individual records cannot be reverse-engineered from statistical reports.

Access Control

Fine-grained permissions control who can access health data and what operations they can perform:

Role	Permissions	Scope	Duration
Primary User	Read all, delete all, export all, grant access	Own data only	Permanent
Primary Caregiver	Read all, receive alerts, view trends	Assigned user data	Revocable by user
Healthcare Provider	Read specified metrics, export reports	Limited to clinical data	Time-limited (90 days default)
Emergency Services	Read emergency context only	Active emergency events	Event duration + 24 hours
Research (Anonymized)	Read aggregated statistics	De-identified population data	Per study protocol

Offline Operation & Synchronization

Seniors may travel to areas with poor connectivity or experience Wi-Fi outages at home. The device must continue functioning during offline periods and seamlessly synchronize when connectivity resumes.

Local Buffering Strategy

// Offline Buffer Management
class OfflineDataBuffer {
  constructor() {
    this.maxBufferSize = 7 * 24 * 60 * 60 * 1000; // 7 days in milliseconds
    this.compressionEnabled = true;
    this.priorityLevels = ['CRITICAL', 'HIGH', 'NORMAL', 'LOW'];
  }

  // Store data locally when offline
  async bufferData(healthData) {
    const entry = {
      timestamp: Date.now(),
      data: healthData,
      priority: this.calculatePriority(healthData),
      compressed: false,
      syncAttempts: 0
    };

    // Compress older data to save space
    if (this.shouldCompress(entry)) {
      entry.data = await this.compress(entry.data);
      entry.compressed = true;
    }

    // Store in local SQLite database
    await this.storage.insert('offline_buffer', entry);

    // Check if buffer is approaching capacity
    if (await this.getBufferSize() > this.maxBufferSize * 0.9) {
      await this.pruneOldData();
    }
  }

  // Prioritize data for synchronization
  calculatePriority(healthData) {
    // Critical: Emergency alerts, severe abnormalities
    if (healthData.alerts && healthData.alerts.some(a => a.severity === 'CRITICAL')) {
      return 'CRITICAL';
    }

    // High: Moderate abnormalities, daily summaries
    if (healthData.alerts && healthData.alerts.length > 0) {
      return 'HIGH';
    }

    // Normal: Regular periodic data
    return 'NORMAL';
  }

  // Sync when connection restored
  async syncWhenOnline() {
    const connectionAvailable = await this.checkConnection();
    if (!connectionAvailable) return;

    // Get all pending data, ordered by priority
    const pendingData = await this.storage.query(`
      SELECT * FROM offline_buffer
      ORDER BY
        CASE priority
          WHEN 'CRITICAL' THEN 1
          WHEN 'HIGH' THEN 2
          WHEN 'NORMAL' THEN 3
          WHEN 'LOW' THEN 4
        END,
        timestamp ASC
    `);

    // Batch upload with retry logic
    const batchSize = 100; // records per request
    for (let i = 0; i < pendingData.length; i += batchSize) {
      const batch = pendingData.slice(i, i + batchSize);

      try {
        await this.uploadBatch(batch);
        await this.storage.delete('offline_buffer', batch.map(b => b.id));
        console.log(`Synced ${batch.length} records`);
      } catch (error) {
        console.error('Sync failed for batch:', error);
        // Mark failed attempts, will retry later
        await this.incrementSyncAttempts(batch);
      }
    }
  }
}

Data Compression Techniques

Minimizing data transmission size reduces battery consumption, cellular data usage, and storage costs. WIA-SENIOR-007 employs multiple compression strategies optimized for time-series health data:

• Delta Encoding: Store differences between consecutive readings rather than absolute values. Heart rate changing from 72 to 73 BPM is encoded as "+1" instead of "73".
• Run-Length Encoding: When values remain stable (common during sleep), encode as "value + duration" rather than repeating the value hundreds of times.
• Quantization: Reduce precision for metrics where high precision isn't clinically meaningful. Temperature can be rounded to 0.1°C instead of 0.01°C without loss of diagnostic value.
• Structured Compression: Use protocol buffers or MessagePack instead of verbose JSON for binary transmission, achieving 60-80% size reduction.