Data

Stream IoT Data to Infuse DB

Move device telemetry into a queryable data store.

Use this guide to design the bridge between Infuse IoT device workflows and Infuse DB ingestion.

Infuse IoT is the device communication and operational context layer. Infuse DB is the scalable time-series data layer for live and historical device records. A production pipeline usually has a backend service or gateway between them.

Integration Shape

---
toolbar:
  title: IoT telemetry to Infuse DB
---
sequenceDiagram
    participant Device as IoT device
    participant IoT as Infuse IoT
    participant Worker as Your ingestion service
    participant DB as Infuse DB

    Device->>IoT: Sends state, route, KV, logger, or telemetry context
    Worker->>IoT: Reads device identity, state, definitions, and route context
    Worker->>Worker: Decodes and validates records
    Worker->>DB: Writes queryable device data
    Worker->>IoT: Optionally queues KV updates or RPC commands

1. Provision The Device

Register and provision the device in Infuse IoT first. The device record gives your pipeline stable identifiers such as deviceId, mcuId, board, organisation, and metadata.

See Connect an IoT Device for the full onboarding sequence.

2. Define The Target Dataset

Create or select the Infuse DB dataset that should hold the device records. The dataset should include the Infuse deviceId and any dimensions you need for querying, such as customer, environment, location, asset group, organisation, board, deployment site, firmware version, or measurement type.

See Datasets and Ingestion.

3. Read IoT Context

Before writing records, your ingestion service can read Infuse IoT context:

Context	Endpoint
Device record	`GET /device/deviceId/{deviceId}`
Device state	`GET /device/deviceId/{deviceId}/state`
Last route	`GET /device/deviceId/{deviceId}/lastRoute`
Logger state	`GET /device/deviceId/{deviceId}/loggerStates`
KV entries	`GET /device/deviceId/{deviceId}/kv/entries`
TDF definitions	`GET /defs/tdf`
KV definitions	`GET /defs/kv`

Use definitions to decode structured payloads consistently. Store the definition version with ingested records when you need reproducible decoding later.

4. Write Records To Infuse DB

Your ingestion service should validate device identity, decode payloads, normalize timestamps, and write raw or computed records into Infuse DB. Keep the raw device identifier and decoded fields together so later queries can join operational data back to the device registry.

Recommended record fields include:

Field	Why it helps
`deviceId`	Joins records to Infuse IoT and Marketplace workflows.
`organisationId`	Preserves account ownership.
`boardId`	Supports hardware-family queries.
`customer`	Supports customer-scoped fleet queries.
`environment`	Separates production, test, staging, or field-trial data.
`location`	Supports site, region, or installation queries.
`assetGroup`	Groups devices by physical asset, product line, or operational unit.
`observedAt`	Captures the device or gateway timestamp.
`ingestedAt`	Captures when the backend wrote the record.
`definitionVersion`	Makes decoding traceable.
`payload`	Stores raw or computed measurement, event, signal, or status data.

5. Use IoT Feedback Loops

After data is ingested, your backend can use Infuse IoT to act on device state:

queue a KV update when configuration needs to change,
send an RPC when an operator or automation requests action,
inspect logger state before downloading more logged data,
read last route before sending downlink work.

For endpoint schemas, see the Infuse IoT API reference and Infuse DB docs.

Data Guides

Task-focused guides for data ingestion and storage workflows.

IoT Guides

Task-focused guides for registering, provisioning, connecting, and verifying Infuse IoT devices.