Skip to main content

WEDA Overview

What is WEDA?​

If we compare an IoT project to running a "Chain Bubble Tea Shop":

  • Your Hardware Devices (IPC/Chips) are the "Branches" scattered everywhere.
  • Your AI Programs or Applications are the "Secret Drink Recipes".

Before WEDA, if you wanted to change the recipe (update software), you had to send someone to every single branch to teach them manually (manual updates). With WEDA (WISE-Edge Developer Architecture), you have a "Central Kitchen Command System". You only need to press one button at headquarters (the Cloud), and the new recipe (Software Container) is automatically and securely sent to 500 branches worldwide, ensuring every drink tastes exactly the same (Standardization).

Define WEDA in one sentence: It is Advantech's "AIoT Scalability Engine". It not only solves hardware-software integration problems but also connects ecosystem resources, helping businesses quickly move from single-point verification to large-scale commercial applications, adapting to the fast-changing AI era.

What Pain Points Do We Solve?​

Anyone who has done an IoT project knows that the hardest part is not writing the code, but what happens after the code is written:

Hardware is like the United Nations; coding is hard (Fragmentation)​

Pain Point: You have cheap chips, powerful AI servers, and old industrial computers at the site. In the past, you had to write different versions of code for each machine, which exhausted engineers.

WEDA's Magic: We use Containerization Technology to "package" the code. Like a shipping container—whether it's on a ship, truck, or plane, the container fits. WEDA allows the same code to run directly on different hardware.

Managing 1 machine is easy; 1,000 is a disaster (Scalability)​

Pain Point: Testing one machine in the office is smooth. But when you have 500 machines scattered across the country, updating software or checking for crashes means running around endlessly.

WEDA's Magic: We provide Cloud Batch Operations. On the cloud dashboard, you can select 500 machines at once and press a button to execute an "OTA (Over-The-Air) Update," just as simple as updating iOS on your phone.

Blind on-site status; debugging is like guessing (Blind Operations)​

Pain Point: A device crashes or AI accuracy drops, but the engineer isn't there and has no idea what happened, so they just guess.

WEDA's Magic: We introduced Digital Twin technology. Every physical device has a "Virtual Twin" in the cloud. You can check this twin's real-time status (temperature, AI confidence level, error codes) anytime, just as clearly as if you were standing right next to the device.

How Does WEDA Work?​

WEDA's logic is "Command from the Cloud, Execute at the Edge". It consists of three main roles:

WEDA Core: The Brain in the Cloud​

  • Role: The Commander.
  • Task: Lives in the cloud server. You use it to view the Digital Twin status of all devices, set permissions, and most importantly, "Send Deployment Commands".
  • Keywords: Command Center, API, Digital Twin.

WEDA Node: The Shop Manager at the Edge​

  • Role: The Executor.
  • Task: A small program living inside every edge device. It listens to the Cloud Commander, downloads software, executes programs, and reports if the device has a fever (overheating) or crashes.
  • Keywords: Edge Agent, Runtime, Keep-alive.

Enablers: The Developer's Arsenal​

  • Role: The Best Teammate.
  • Task: It packages difficult hardware drivers for you. Want to grab data? Want to run AI? Just call the existing functions; no need to code from scratch.
  • Keywords: SDK, Library, Drivers.

What Can You Actually Do with It?​

WEDA is not picky. As long as it is software running on edge devices, WEDA can manage it:

  • AI Defect Inspection in Factories:
    • You train an AI model to identify product defects.
    • Use WEDA to deploy it to 50 computers on the production line with one click.
    • Advanced Use: Through Digital Twin, you can monitor the GPU temperature and AI inference speed of every computer in real-time. If it overheats, the cloud alerts you immediately.
  • Traffic Monitoring in Smart Cities:
    • Cameras at intersections count traffic flow.
    • Advanced Use: AI becomes inaccurate on rainy days? No problem. From your office, use WEDA to remotely push a new "Rainy Day Model" to solve the problem instantly.
  • Data Monitoring for Solar Panels (Energy Management):
    • You write a Python program to collect power generation data.
    • Use WEDA to deploy it to gateways at various sites.
    • Internet disconnected? No worries. WEDA Node will temporarily save the data for you and automatically upload it when the connection is back.

How to Start?​

Just like buying software, we offer two modes:

  1. SaaS Public Cloud (Like subscribing to Netflix):
    • Suitable for: Testing the waters, or if you don't want to manage servers yourself.
    • Advantage: Sign up and use immediately. Start developing right away.
  2. Private Cloud (Like buying a DVD to watch at home):
    • Suitable for: Large enterprises, data that cannot leave the premises, or if you want full control.
    • Advantage: The entire system is installed on your company's intranet for maximum security.

Cheat Sheet for Beginners​

TermExplanation in Plain English
ContainerA technology to package programs (like Docker). It guarantees the code runs on your computer and at the site.
Container StackA standardized application package based on Docker Compose. It defines exactly which containers to run and their configurations (essentially your docker-compose.yml file).
Digital TwinA "Twin" of the device in the cloud. You can know the device's current status without going to the site.
OTA (Over-The-Air)Updating software wirelessly over the internet, without plugging in a USB cable.
TelemetryHealth check reports actively sent back by the device (e.g., "My CPU temperature is 60 degrees").
WEDA CoreThe central management platform in the cloud, responsible for coordinating all devices, applications, and data.
WEDA NodeThe agent program on the edge device; it is the executor of cloud commands.

What are the Next Steps?​

It depends on what kind of developer you are:

  • I am an Edge Application Developer:
    • Do you want to handle hardware control or write Python/C# apps?
    • Please go to [WEDA Node & Enablers Development Guide].
  • I am a System Integration Developer:
    • Do you want to retrieve device data for the company ERP or build a dashboard?
    • Please go to [WEDA Core Integration & Enablers Development Guide].

(Note: Whether you use a cheap single chip or a high-end AI Server costing hundreds of thousands, WEDA supports them all!)

On this page ...

Is this helpful?