PR97841

NEW TAIPEI CITY, Taiwan, Sept. 22, 2022 /PRNewswire=KYODO JBN/ --

Aetina (http://www.aetina.com/), a leading edge AI solution provider,

introduced its end-to-end AI management solution in an NVIDIA GTC on-demand

session

(https://www.nvidia.com/gtc/session-catalog/?search=A41137&tab.catalogallsession

stab=16566177511100015Kus#/) this week. The solution helps global AI partners

and clients of Aetina successfully adopt edge AI using NVIDIA AI development

and deployment tools, as well as Aetina's NVIDIA AI-powered training and

inference platforms.

In the session, Jeremy Pan, AI Solutions Product Manager at Aetina, deep dived

into Aetina's solution, and shared an impactful application case of AI-powered

automatic optical inspection (AOI) in a client's factory in Yilan, Taiwan.

Challenges of Edge AI Adoption

To adopt edge AI, system integrators and developers need to train AI models and

deploy them on edge devices. The AI model training process involves collecting

and labeling large amounts of data using high-performance computing platforms,

which can result in high training costs. These factors make the training

process challenging and time consuming for system integrators and developers.

AI model deployment can also be difficult when the system integrators and

developers have multiple remote edge devices in different locations. The

challenges of AI model deployment include optimizing models for inference to

run efficiently on edge devices, long deployment processes, difficulties of

team communication and collaboration, high follow-up monitoring costs caused by

large number of AI models, security issues, and high costs.

End-to-End AI Management Solution

Aetina provides an end-to-end solution to help its partners and clients develop

and deploy AI applications for the edge. The solution consists of Aetina's

NVIDIA-Certified

(https://www.nvidia.com/en-us/data-center/products/certified-systems/)edge

computing platforms and NVIDIA's AI model development and deployment tools.

These tools include NVIDIA Fleet Command™, and the NVIDIA AI Enterprise

(https://www.nvidia.com/en-us/data-center/products/ai-enterprise/) software

suite, which provides enterprise support for the NVIDIA TAO

(https://developer.nvidia.com/tao)toolkit and NVIDIA Triton Inference Server™

(https://developer.nvidia.com/nvidia-triton-inference-server).

With Aetina's NVIDIA-Certified edge computing platforms and other NVIDIA tools,

Aetina streamlines the AI training and deployment process, helping partners and

clients adopt edge AI more quickly.

Ongoing Application Case in Smart Factory

One of Aetina's clients, a global provider of industrial embedded flash and

DRAM solutions, planned to add AI-powered AOI systems in its factory for better

productivity. The flash and DRAM products that Aetina's client produces are

small and complex electronic components designed for harsh environments and

applications; the producer of these components needed an AOI system capable of

processing high-resolution image recognition tasks with high processing speed.

Aetina helped the client develop a prototype of the AI-powered AOI system.

An Aetina solution team first trained an AI model capable of finding defective

products in factory production lines for the AOI system with NVIDIA TAO on

Aetina's AI training platform—SuperEdge

(https://www.aetina.com/products-features.php?t=335) AIP-D422—and then uploaded

the model to NVIDIA GPU Cloud(https://www.nvidia.com/en-us/gpu-cloud/) (NGC).

With NVIDIA Fleet Command™, the solution team remotely deployed the model on

Aetina's AI inference platform—MegaEdge

(https://www.aetina.com/products-features.php?t=336)AIP-FQ47—in the factory of

Aetina's client from NGC, successfully developing the prototype of the AOI

system.

With AOI cameras, the prototype of the system is now able to greatly speed up

the inspection task of finding the DRAM products with incorrectly placed

corporate stickers with high accuracy in less than one second, compared to a

human inspector that usually takes about ten seconds to do the same task. When

the AI-powered AOI system is fully built in the future, it will be installed in

the factory of Aetina's client to run inspection tasks in multiple production

lines.

In addition, NVIDIA Fleet Command™ allows the Aetina solution team to quickly

update the AI model on edge devices in the factory production lines of Aetina's

client, enabling efficient remote AI management—a key to successful edge AI

deployment.

The end-to-end AI management solution is a part of Aetina Pro-AI Service—which

helps global partners and clients adopt AI for different vertical applications

besides AOI in factories, with Aetina's edge AI hardware and software.

Source: Aetina Corporation

Image Attachments Links:

   Link: http://asianetnews.net/view-attachment?attach-id=429491

   Caption: Aetina Pro-AI Service—a service to help global partners and clients of Aetina

adopt AI for different vertical applications with Aetina’s edge AI hardware and

software.