WEARABLES & IOT CONNECTORS

Integration Architecture

3 of 7

Choose the right path to connect devices and IoT platforms to Mediloop through BLE, mobile health frameworks, provider clouds, gateways or controlled local networks with resilient time-series ingestion.

Integration FlowsComponentsProtocols & SecurityDevice OnboardingData IngestionReliability & OfflineScalability

Integration Flows

Direct to Mobile App

BLE / GATT

Local pairing, permissions, real-time/periodic sync and offline buffering.

Platform Integration

HealthKit / Health Connect

Native platform APIs, user consent and standardized mobile health records.

Manufacturer Cloud

OAuth / API / Webhook

Provider authorization, API polling and/or event delivery.

Home Gateway / IoT

Wi-Fi / Ethernet / MQTT

Gateway buffering, reliable transmission and multiple local devices.

Clinical / Local Network

Workstation Agent / local gateway

Controlled local-network integration for specialized clinical devices.

Overall Architecture

Devices & Data Sources

Wearables, medical devices, IoT sensors and provider clouds

Connectivity Layer

BLE, platform APIs, OAuth/REST, webhooks, MQTT and local gateway

Mediloop Ingestion

API gateway, onboarding, validation, normalization and event processing

MSIS – Semantic Layer

FHIR mapping, terminology, identity reconciliation, quality and context

Mediloop Platform

Patient record, trends, alerts, RPM and EHR/HIS integration

Key Components

Device Registry

Device identity, model, capabilities, ownership and operational state.

Pairing & Consent

User authorization, provider connection and permission lifecycle.

Ingestion Services

REST/webhook/MQTT ingestion plus local/gateway upload.

Normalization Service

Map to FHIR, terminology and UCUM while preserving source values.

Time-Series Store

Scalable storage/stream processing for high-frequency longitudinal data.

Device Management

Connection health, provider token lifecycle, firmware metadata and disconnect/decommission.

Protocols & Standards

Bluetooth LE / GATT

Apple HealthKit

Android Health Connect

OAuth 2.x / OpenID Connect where applicable

HTTPS REST APIs and webhooks

MQTT over TLS for applicable gateways/IoT

FHIR Device / DeviceMetric / Observation / Provenance

IEEE 11073 where applicable

LOINC / SNOMED CT / UCUM through semantic mapping

Local gateway / Workstation Agent for controlled clinical environments

Data Flow Example

Device generates measurement

Heart rate, glucose, activity or other sensor data

App / cloud / gateway receives

BLE, HealthKit/Health Connect, provider API or MQTT

Mediloop validates

Authenticate source, schema validation and patient/device linkage

MSIS normalizes

Map codes/units, add context, deduplicate and preserve provenance

Mediloop uses data

Record, trends, alerts, dashboards and clinical workflows

jsonCopy

{
  "deviceId": "garmin-12345",
  "type": "heart-rate",
  "timestamp": "2026-08-27T10:15:30Z",
  "value": 72,
  "unit": "bpm",
  "source": "provider-cloud",
  "quality": "good"
}

Developer Integration Surface

Surface	Operation / resource	Use	Status
BLE / GATT	Discover, pair, read/write, notifications	Direct mobile/local device path	Protocol surface
HealthKit / Health Connect	Native platform read/write/query APIs	Mobile ecosystem integration	External platform API
Provider cloud	OAuth + REST/webhooks per provider	Manufacturer cloud path	External provider API
MQTT / HTTPS	Publish/subscribe or upload to gateway/ingestion	Home gateway and IoT path	Protocol surface
FHIR	Device, DeviceMetric, Observation, Provenance	Canonical clinical/interoperable representation	FHIR surface
Mediloop Connector SDK	Adapter, parser, mapper and event handler APIs	Custom provider or device connector development	SDK contract evolving

Reliability & Offline

Use local buffering for intermittent BLE/network connectivity.

Keep original device timestamps and sequence identifiers when available.

Deduplicate retries and replayed provider events.

Use idempotency/correlation identifiers for cloud/gateway ingestion.

Detect missing samples and data gaps rather than silently interpolating.

Define retry/backoff and dead-letter handling for provider/gateway failures.

Best Practices

Use standard protocols and FHIR resources where they fit.

Request the minimum permissions necessary for the use case.

Normalize incoming data without discarding original source values.

Handle device clocks, time zones and offline synchronization explicitly.

Monitor device/provider health separately from data-quality status.

Keep regulated clinical uses distinct from consumer-wellness scenarios.

Common Challenges

Provider-specific formats

Use adapter-specific parsing plus canonical FHIR/MSIS mapping.

Clock drift & irregular samples

Normalize to UTC while preserving original timestamps and quality flags.

Duplicate / missing data

Deduplicate by source identifiers and detect expected-vs-received samples.

Connectivity limitations

Buffer locally, retry safely and expose connection health.

Regulatory context

Track intended use, consent and data-source classification.

Next Steps

Choose path

BLE, mobile platform, cloud, gateway or local clinical network

Define identity, patient linking and connection metadata

Map & validate

FHIR, terminology, units, provenance and quality

Test

Representative devices, offline cases, retries and data gaps

Production

Monitoring, provider token lifecycle and support ownership