How does Avnet help build a big data management and analysis system?
In the era of Internet of Things (IoT) and Industry 4.0, the demand for massive data and real-time processing is growing, posing unprecedented challenges to edge computing devices. In response, Avnet and NXP have launched a series of "out-of-the-box" artificial intelligence solutions based on the innovative MCX-N series microcontrollers, aiming to provide customers with efficient, reliable and flexible artificial intelligence services, accelerate the application deployment of machine learning (ML), and promote the development of edge intelligence.
Hardware advantages:
Provide strong support for IoT and industrial automation
Dual-core system improves energy efficiency and reduces power consumption
The MCX N series of industrial and Internet of Things (IIoT) MCUs include the MCX N94x and MCX N54x models, both of which are based on dual high-performance Arm ® Cortex ® -M33 cores running at 150MHz, with 2MB flash (optional full ECC RAM), DSP coprocessor and integrated eIQ Neutron NPU neural processing unit. Compared with a separate CPU core, the machine learning throughput with NPU can be increased by up to 42 times, which can reduce wake-up time and reduce overall power consumption.
Integrating PowerQuad and Coolflux BSP32
The powerful DSP coprocessor PowerQuad can greatly improve the execution performance of common DSP algorithms in MCU. In the field of power electronics applications, the FFT/iFFT, FIR, and IIR modules included in PowerQuad can provide a more efficient solution than the Arm Cortex-M core.
NXP's self-developed audio processing unit Coolflux BSP32 coprocessor is used for audio and voice processing. Combined with the core and PowerQuad, it provides a new solution for low-power and high-performance edge artificial intelligence applications.
Advanced Information Security
Built according to NXP’s “Secure by Design” methodology, it provides secure boot with an immutable root of trust and hardware-accelerated encryption, as well as a built-in EdgeLock® secure enclave Core Profile, enabling products to be designed to meet industry standards.
Wide range of peripherals
This series of products can provide engineers with the flexibility needed for innovative designs and can be customized and optimized according to different application requirements. Among them, the MCX-N94x series is aimed at industrial applications and has a wider range of analog and motor control peripherals. The MCX-N54x series is aimed at consumer electronics and IoT applications and has a variety of peripherals, including high-speed USB with PHY, SD card, and smart card interface.
Software advantages:
Accelerate development and enable innovation
To support customer development, Avnet provides a wealth of MCUXpresso developer tools, including IDE, SDK and security configuration tools, to help developers simplify the development process and improve development efficiency. In addition, the solution also integrates a variety of advanced artificial intelligence algorithms, which can achieve accurate recognition and processing of various data types such as images, voices, and texts, and is compatible with various artificial intelligence algorithms and models to meet the personalized needs of different applications.
Application scenario: AFCI detection of solar inverters and circuit breakers
AFCI can identify and cut off the arc as soon as it occurs to ensure the safety of the distributed photovoltaic system and avoid fire caused by high temperature of the arc.
Application scenario: Industrial anomaly detection
Click above to watch the video ????
This video shows a simple fan anomaly detection demo, which uses 3-axis accelerometer data at a sampling rate of 200Hz. The system uses a support vector machine (SVM) algorithm optimized for MCU, allowing users to redefine what is a "normal" state as needed. By calculating features such as RMS, dominant FFT frequency, dominant FFT amplitude, 256-point FFT, and 100-point sliding window standard deviation in real time, the system can complete data collection and model training in just 30 seconds. And perform real-time health value assessment on the monitored equipment, and determine whether the equipment is operating normally based on this assessment.
Application scenario: Person/face detection
Click above to watch the video ????
The video shows the solution-based personnel detection, multi-face detection and face tracking detection demos. All demos use the innovative micro YOLOv3 style network architecture. The LCD screen can display the detected faces in real time. Due to the use of NPU acceleration, the face detection and multi-face detection time is 24ms, while the face tracking detection time is 21ms.
The MCX-N series of microcontrollers and their supporting artificial intelligence solutions have great potential in improving the intelligence level of equipment, optimizing resource utilization efficiency, and enhancing system security. In the future, Avnet and NXP will continue to work closely together to jointly develop more innovative technologies and accelerate the implementation of edge artificial intelligence.
Interactive topics
▍The interactive topic of this issue is: How does artificial intelligence reshape the industrial automation landscape? Talk about your views and prospects for its future ~
You are welcome to participate in the interaction at the end of the article, any of which is fine [Like/Reading/Comment]. Users who click "Reading" and "Like" need to send the screenshot to the backend of the official account. We will randomly select 2 users from all participating users to give away 1 Avnet customized charging cable.
▍Last list announced
The winners and prizes of the article in issue 240725, "Digitalization of Industrial Operations: Low Power Consumption, High Performance, and Super Connectivity Are the "Superpowers"!" are as follows:
اGive you a millionاCute baby, happy little bunny
Customized charging cable
*Please the above winners click here to fill in the prize redemption information. The staff will arrange the gift distribution as soon as possible after verification.
*This event is open only to people in the electronics industry. Participating in this event means you agree to Avnet's privacy policy.
*Avnet reserves the right of final interpretation of the event.
京公网安备 11010802033920号