Documentation

This documentation explains the Jiwu AIoT Platform module by module: creating a product, onboarding devices, reading data, building data screens, using the AI Assistant, opening the API and deploying on premises. Read it in order, or jump straight to the chapter you need.

3 sections · Updated with each platform release

How to read this documentation

New to the platform: read Products → Devices → Data Screens in that order; this path takes one data point from the device all the way to a screen.
Integrating your own systems: go straight to "Credentials and OpenAPI" and "Protocol gateways".
Deploying to your own data center: read "On-premises deployment".
The "Note" at the end of every chapter describes a pitfall seen in real projects — do not skip it.

The full path of one data point

Understanding this path explains how the modules divide the work, and tells you which module to look at when something goes wrong.

Create a product in "Products" and define which properties it has (for example temperature, pressure, switch state).
Register the physical device under that product in "Devices"; the device now has a platform identity.
The device reports over MQTT, or a Modbus gateway polls it, and data is stored according to the product and its thing model.
Inspect the live data stream and historical curve in device details; send a downlink command when you need to control it.
In "Data Screens", drag the metrics you care about into a dashboard, bind device data and publish it.
In the "AI Assistant", query this data in natural language, or schedule a recurring inspection job.
When an external system needs the data, create a credential in "Credentials" and read it through the OpenAPI.

NoteData normally reaches the platform within seconds. Data screens refresh on an interval (10 seconds by default, 5 seconds minimum), so do not use a screen to judge millisecond-level timing.

Module quick reference

Products: define "what this kind of device looks like" — model once, reuse in bulk.
Devices: manage "this specific device", including registration, status, data and commands.
Data Screens: turn data into a picture you can show others; publishable and shareable.
AI Assistant: query data, send commands and run scheduled inspections in natural language.
Credentials and OpenAPI: give third-party systems a controllable, revocable key.
Protocol gateways: how devices deliver their data (MQTT direct or Modbus polling).
On-premises deployment: deliver the whole platform to your own servers.

Chapters

7 chapters, grouped by module.

Products

A product is the most basic concept in the platform: it describes what one kind of device looks like. Define the properties, events and services of similar devices once in a product, and every device you onboard afterwards reuses that definition instead of being configured one by one.

5 sections →

Devices

Device management is about "this specific device". It is where devices are registered, monitored day to day, inspected for data, sent commands and tagged with a location — so it is the module operations staff spend most time in.

6 sections →

Data Screens

Data screens turn device data into something you can put on a wall: drag components onto a canvas, bind device data and publish a shareable dashboard. No frontend code and no separate deployment are needed.

6 sections →

AI Assistant

The AI Assistant is not a standalone chat tool but an agent that operates the platform itself: it wraps devices, products, commands, time-series data, the system and users into tools. You describe a goal in natural language and it picks tools, reads data, gives conclusions and sends commands when needed.