In the food industry, quality inspection and classification of raw materials (such as grains, legumes, fruits and vegetables) are crucial for ensuring product quality. Modern food factories commonly employ sorting machines equipped with machine vision technology to classify raw materials. However, this process still faces numerous challenges, such as the impact of ambient light variations on image interpretation, the limited ability of traditional machine vision systems to detect minor defects and the need for system stability during high-speed operations.
Traditional machine vision technology primarily relies on predefined algorithms and optical principles for inspection. However, when raw materials exhibit color variations or surface defects due to natural factors, the error rate of machine vision systems increases significantly. Moreover, food ingredient sorting machines must analyze images and make classification decisions at extremely high speeds. Conventional image processing technologies are constrained by computing power and response speed, often failing to meet real-time processing demands.
To address these challenges, the industry has begun integrating edge AI technology for real-time image recognition, enhancing sorting accuracy and efficiency.
To overcome technical bottlenecks in raw material sorting, sorting machines are now integrated with edge AI systems to enhance image analysis accuracy and real-time performance. Among these, the PJAI-100-ON serves as a fanless edge AI system powered by the NVIDIA Jetson Orin module, delivering up to 100 TOPS of AI computing power. This enables the system to process large volumes of high-resolution image data, ensuring the sorting equipment can quickly analyze raw material characteristics and classify them with precision.
In food ingredient sorting applications, the PJAI-100-ON can instantly detect surface defects, foreign material contamination or color deviations in raw materials. With a response time of less than 1ms, it guarantees real-time classification decisions. Given the high precision required for raw material inspection, this system employs AI-driven defect detection technology, which continuously refines classification standards based on historical learning data, improving detection accuracy and reducing false positives.
Once the classification is determined, the edge AI system communicates with a PLC to control the sorting mechanism through valve operations. Additionally, the edge AI system processes image data locally, preventing latency issues and ensuring production line efficiency, allowing food factories to execute quality sorting operations with greater reliability.
Designed specifically for Edge AI applications, the PJAI-100-ON is an industrial computing system that not only offers ample AI computing power but is also engineered for industrial environments. Its fanless design ensures stable operation even in high-dust food production environments. The system is equipped with 4GB/8GB LPDDR5 memory for high-speed image data processing and features an M.2 2280 M-Key NVMe (PCIe x4) slot, supporting high-speed storage expansion for enhanced system scalability.
Additionally, the PJAI-100-ON is equipped with two RJ45 GbE network ports, ensuring stable network connectivity and seamless integration with industrial cameras.
For I/O interfaces, the device offers three USB 3.2 Type-A ports and two DB9 COM ports (supporting RS-232/422/485 communication), providing flexible connectivity options for various sensors. This enables food manufacturers to configure their hardware according to operational needs, improving system integration and scalability.
With its 100 TOPS AI computing power, sub-1ms data transmission latency, and flexible I/O expansion, the PJAI-100-ON delivers exceptional performance and stability in food ingredient sorting applications. This solution not only enhances quality inspection accuracy but also optimizes production efficiency and process reliability, offering food factories a more intelligent edge AI application for automated sorting.
Portwell provides a comprehensive DMS (Design, Manufacturing, Service) solution, ensuring customers can rapidly deploy and optimize application solutions. With a dedicated R&D team, Portwell delivers customized hardware and firmware designs tailored to specific application needs, ensuring solution optimization.
The high-quality manufacturing process guarantees stable and reliable equipment performance, even in harsh industrial environments. Moreover, comprehensive technical support includes remote diagnostics, system maintenance and software updates, ensuring long-term operational efficiency.
Through DMS services, Portwell not only provides hardware solutions but also empowers customers with efficient, reliable and flexible intelligent manufacturing applications, helping them gain a competitive edge in the food industry.