hero-image

NEWS

AI-Powered Object Counting Application To Adopt Cloud Services
calendar
2022.03.02
repeat
2023.12.29

AI-Powered Object Counting Application To Adopt Cloud Services

About project

This application development project started in 2021 with a view to applying Artificial Intelligence (AI) into manufacturing activities to automate workers’ task, shorten production time and enhance management.

During our client’s production, there is a stage where the workers have to directly count and identify the exact number of minuscule objects limited in a small area. Each object’s size is only 1 to 2 millimeters, and there can be hundreds of objects to be counted each time. The client expected that Kaopiz’s Image Recognition AI solution would automate and shorten the time spent on such object counting progress once this project is completed. By deploying and operating Kaopiz’s AI engine on the cloud, the workers can use a smartphone to snap photos of the objects, send them to the AI engine and let it automatically count and list up the object numbers with only insignificant errors. Aside from that, other functions such as reporting, account management, number of counting time statistics management also smoothen the business flow.

Challenges

The application has been developed in this project allows users to take photos using camera on their smartphones and send them to the cloud. The Computer Vision and Artificial Intelligence program will then detect, count the insects objects and mark the detected objects on the snapshot.

The client has many strict requirements for this new system: fast processing speed identified as the top priority; ability to handle in case the number of concurrent users increases; storage of all necessary user data with high loading speed; self-defense against unauthorized access and network attacks; effective cost balance.

Solutions

AWS-infra-solution

Kaopiz is an IT outsourcing company based in Hanoi and Tokyo, aiming to be one of the top outsourcing companies in Vietnam. Customer satisfaction is always our top concern. In this project, we need to build a solution with high availability and performance that could handle concurrently multiple requests using complex machine learning models individually developed and customized by Kaopiz engineers.

This application is resource-intensive and compute-intensive because it runs machine learning models. We incorporated Amazon EC2 C5 Instances to deliver efficient cost performance that accelerates machine learning operations such as convolution or deep learning workloads.

In addition, these AWS services were combined with the Auto Scaling Group to deal with sudden fluctuations in the number of active users. Full automation freed the entire resource group from constant monitoring and manual configurations.

Files and data need to be stored in large amount, providing stable, long-lasting, globally accessible retrieval. We chose Amazon Simple Storage Service (Amazon S3) and Amazon Relational Database Service (RDS) for high performance, and rapid recovery from failure.

On the outermost layer, to protect the application and APIs from security breaches and resource overconsumption, we put AWS WAF for traffic control to increase automation and maintain security rules.

Benefits

With project deployment on AWS, we don't need to worry about server system configuration or management. Our problems were completely solved as AWS provides favorable environment to develop and operate applications (especially applications incorporating machine learning or automatic resources’ scalability). Since going live, the system has been in fairly stable condition without experiencing any downtime.

Whenever we receive new change requests or update from customer, we are able to carry out smoothly without any significant disadvantage.

Reasonable cost is also another major advantage of AWS when compared to other cloud services, as well as to renting physical servers. AWS helped the team to minimize the effort spent on server configuration and managment, and to focus their resources on other important goals - user acquisition and product marketing.

Are you considering AWS cloud services? Kaopiz would like to give you a hand!

Being one of few companies in Vietnam joining AWS Partner Network and becoming AWS Consulting Partner, Kaopiz provides a wide range of AWS-related supporting service. We have years of experience in working with various types of systems, from business system, mobile applications to cutting-edge technologies such as Artificial Intelligence and Blockchain, etc. Our experienced AWS engineers can support you from system migration, infrastructure construction to operation and maintenance.

If you have any trouble dealing with troubles related to cloud infrastructure, or if you consider adopting AWS services but don’t have enough human resources, don’t hesitate to contact us. Our sales team will give you a tailored proposal after a discussion in detail with you.

よく読まれている記事

https://stg.kaopiz.com/wp-content/uploads/2023/12/converting-handwritten-characters-into-data-OCR-engine.png
ブログ
23.12.28
手書き文字データ化 | 方法とメリット・デメリットについて徹底解説
手書き文字データ化 の手法とOCR技術の活用について詳しく解説。さらに、OCR技術の種類と、メリットとデメリットについても深掘りしていきます。手書き文字のデータ化に関する具体的な手法が理解でき、どの手法が担当業務に最適かを判断するための知識が手に入ります。
https://stg.kaopiz.com/wp-content/uploads/2023/12/オフショア開発の注意点.png
ブログ
23.12.25
オフショア開発の注意点はこれ!日本企業が納得しておきたいポイントを解説
オフショア開発 には事前に知っておきたい 注意点 があります。日本企業は、これに納得しておかないと、スムーズに開発できないかもしれません。事前に抑えておくべき注意点を解説します。今回は、依頼前の納得感に欠ける方へ向けて、オフショア開発で納得しておきたいポイントについて解説します。
https://stg.kaopiz.com/wp-content/uploads/2023/12/2024-trendy-develop-languages.jpg
ブログ
23.12.21
プログラミング言語とは|2024年に学ぶべきコンピュータ言語12選
決定する際には、自身が取り組もうとしている難易度、すでに持っている プログラミング言語 の知識と既存のコーディングスキルとの合致性、あるいは人気の プログラミング言語 を学びたい理由など、以下では、2023年に雇用者の間で需要が最も高いプログラミング言語について紹介します。