The trend towards decentralised AI and edge computing platforms in industry, automation and critical infrastructure is simultaneously raising the bar in terms of the safety, reliability and resilience of the hardware used. Against this backdrop, Neousys Technology has announced that the embedded systems of the POC-766AWP and POC-900 series have been successfully certified to UL 62368-1. The Taiwanese company is thus underlining its position in the field of robust industrial and edge AI platforms.
Certification to UL 62368-1 is regarded as an important international safety standard for audio, video, information and communication technology. It is increasingly replacing older safety standards and is aimed in particular at modern IT and network environments with high demands on operational safety, thermal stability and protective mechanisms.
Focus on harsh industrial environments
Both the POC-766AWP and the POC-900 series have been specifically developed for use in demanding environmental conditions. According to the manufacturer, both platforms operate within an extended temperature range of -25 °C to 70 °C and withstand shocks and vibrations in accordance with MIL-STD-810H. This makes the systems particularly suitable for industrial automation, vehicle-based applications, outdoor installations or edge AI scenarios in production environments.
The compact POC-766AWP features an IP67-certified housing and has been designed for applications such as AI-assisted visual inspection, real-time analytics and industrial data acquisition. At the heart of the system is an Intel Core i3-N305 processor with eight efficiency cores and DDR5 RAM. It also features a wide input voltage range of 8 to 35 V DC and ignition control functions, making the system suitable for mobile or vehicle-mounted installations.
Meanwhile, the POC-900 Series targets modern AI and automation applications with AMD’s Ryzen PRO 8640U processor, including an integrated Ryzen AI NPU. According to Neousys, the platform is said to achieve up to 2.7 times higher CPU performance compared to the previous generation. At the same time, the integrated AI acceleration enables enhanced local inference processes directly at the edge – without a permanent cloud connection.
Edge AI and machine vision drive requirements
Compact AI-capable systems are becoming increasingly important, particularly in the fields of machine vision, industrial quality control and automated manufacturing. The POC-900 series offers a correspondingly comprehensive range of interface options: four GbE PoE+ ports, four USB 3.2 interfaces and isolated digital inputs and outputs allow for the direct integration of cameras, sensors and control technology.
In addition, Neousys relies on its proprietary MezIO expansion interface, through which further LAN, COM or ignition control modules can be integrated. This modular approach reflects a general trend in industrial edge computing: systems should be adaptable as flexibly as possible to different production, safety or automation environments.
Certifications are becoming a competitive factor
As industrial infrastructures become increasingly networked, not only are performance requirements growing, but so too are regulatory and safety expectations. Certifications such as UL 62368-1 are therefore increasingly becoming a competitive factor – particularly for applications in critical environments or international projects.
Neil Liu, Product Manager at Neousys, emphasises in this context that the certification confirms the company’s commitment to high quality and safety standards. For operators of critical systems, this is particularly important in the case of long-term and safety-critical deployments.
The new certifications complement the company’s existing UL-certified platforms, including the Nuvo-9000 series edge AI systems and the GPU-based AI computer Nuvo-9166GC.
Security sector benefits from robust edge platforms
Robust edge systems are becoming increasingly strategically important for the security sector. Modern video analytics, AI-assisted object recognition, autonomous sensor technology and decentralised security architectures require powerful computing platforms directly on-site – often under adverse environmental conditions.
Particularly in KRITIS environments, transport infrastructure, energy facilities or industrial sites, there is a growing need for compact, certified and fail-safe edge AI systems that can process real-time data even without a permanent cloud connection. Systems such as the new Neousys platforms thus exemplify how traditional industrial IT is increasingly intertwining with the requirements of modern security and AI infrastructures.
