Even developers without ML knowledge can create optimal ML libraries with minimal data
Library, applicable to all STM32 MCUs… “Many applications to electric vehicle platforms”
[Editor's Note] As the use of artificial intelligence becomes widespread across industries, the need for solutions that simplify the development of machine learning applications is increasing. This is because while large corporations typically hire data scientists to collect large amounts of data over several months to create AI models, smaller companies face the reality that it is not easy to even hire data scientists. In response, a new solution has emerged: NanoEdge AI Studio, a utility for embedded developers without data engineering expertise. We met with Hyunsoo Moon, Manager at STMicroelectronics, to hear about NanoEdge AI Studio.
▲ Moon Hyun-soo, Manager of ST Microelectronics (Photo: ST)
I am in charge of MCU technical support at STMicroelectronics (ST) Korea. I am responsible for providing technical support for customers using STM8/STM32 MCU products and for artificial intelligence solutions based on STM32 MCU.
■Although AI is a hot topic in the industry, it seems difficult for companies with limited budgets or no experience to introduce and utilize it. What are the difficulties that you are experiencing in the field? In the past, high-performance hardware systems were required for artificial intelligence processing, so they were implemented through server or cloud environments, which required hardware complexity and high costs.
However, recently, frameworks such as TensorFlow Lite have been developed to support pre-trained neural network processing in MCU-based embedded system environments, making it possible to easily apply neural network models to MCUs and perform inference in actual applications.
However, there are several challenges to developing MCU-based Edge AI devices. Some are related to technology shortages, while others are purely technical or financial. All of these combined make the project too complex to solve, which can make applying AI to MCU-based environments a concern or a high barrier to entry.
The process for collecting high-quality data also requires a lot of effort, but the biggest problem is the lack of development personnel with the capabilities necessary for AI development in the existing MCU-based embedded system development environment.
■What features does ST’s NanoEdge AI Studio provide? NanoEdge AI Studio is a new machine learning (ML) technology that makes it easy for developers using STM32 MCUs to achieve true innovation. NanoEdge AI Studio is an intuitive software tool that allows system designers using ARM-based low-power microcontrollers to easily and quickly apply machine learning algorithms to a variety of applications, including connected products, home appliances, and industrial equipment. Even without any skills or knowledge related to machine learning, the GUI-based NanoEdge AI Studio allows developers to create the optimal ML library for their project based on minimal data in just a few steps. NanoEdge AI Studio can create four types of libraries: △Anomaly Detection △Outlier Detection △Classification △Regression.
▲ NanoEdge AI Studio (Image-ST)
One of the great things about NanoEdge AI Studio is that it doesn’t require any specific data science skills. Any software developer using NanoEdge AI Studio can create optimal ML libraries in a user-friendly environment with absolutely no artificial intelligence (AI) skills.
Since ML library creation is possible based on Arm® Cortex®-M0/M0+/M3/M4/M7, ML libraries created through NanoEdge AI Studio can be used on all STM32 MCUs. It is also possible to create highly accurate ML libraries even with a very small amount of datasets.
Developers using STM32 MCUs can automatically generate a C code-based ML library by simply entering the required dataset into NanoEdge AI Studio.
■What are some specific application cases of NanoEdge AI Studio? Recently, the use of NanoEdge AI machine learning libraries on electric vehicle platforms is increasing. Prediction and monitoring libraries are being used based on a large amount of data that can be logged on electric vehicle platforms, such as torque prediction and stator winding temperature prediction in electric vehicles. NanoEdge AI's machine learning libraries are being applied in various fields, such as machine learning libraries for human behavior recognition in smartphones and energy consumption prediction in the energy field.
In addition, NanoEdge AI's machine learning algorithms are being used in many fields such as △behavioral recognition △chemistry △energy △healthcare △industry △smart home, etc.
■Why is Edge AI necessary in this day and age, and what is your outlook for the future? Recently, MCU is the core of edge devices, and MCU is always used in the system at the end. Therefore, the inclusion of AI functions in this area enables improved application implementation in various areas such as powerful performance and increased efficiency.
It provides a new user experience to users while improving the existing algorithm method. Since input data such as sensor data can be inferred using an AI-based algorithm method in the edge device area that includes the MCU, the cost required for server or cloud connection is reduced. Additionally, since it does not require network connection, it can maintain high stability in terms of security.
In this way, the demand for AI-based processing and combination with IoT platforms in the edge device area is expected to increase very rapidly.
■What differentiated solutions does NanoEdge AI Studio plan to expand its scope in the future? NanoEdge AI Studio is being prepared so that developers using STM32 MCUs can create highly accurate ML libraries without additional knowledge or technical skills in artificial intelligence, and the process of creating these ML libraries can be done simply and easily. In order to create highly accurate ML libraries with a small amount of datasets, various ML algorithms will be developed and improvements to existing ML algorithms will continue to be updated.
■Lastly, please say a word to e4ds readers. ST is providing NanoEdge AI Studio, which allows developers familiar with embedded development environments to easily apply AI solutions on products. ST will continue to update AI-related solutions and demo distributions in the future. As the importance of edge devices increases, ST will be a satisfactory alternative for building a platform that enables inference based on AI models in the On-Device format.
thank you
좋은 내용이네요