Aetina launches Mini Series edge AI systems for industry

Aetina has launched a Mini Series of edge AI systems for industrial vision AI and generative AI deployments. The range includes four compact models built around NVIDIA Jetson Orin modules.

The line-up consists of the AIE-CO23/33-S1, AIE-CN33/43-A1, AIB-MO23/33-S1 and AIB-MN33/43-S1. The systems are aimed at industrial edge environments where space is limited and local processing is needed for visual data.

They are designed for users moving AI projects from proof-of-concept work into operational deployments, including enterprises and system integrators that need systems to connect directly to cameras and sensors and operate reliably in demanding environments.

Compact designs

Two models, the AIE-CO23/33-S1 and AIE-CN33/43-A1, use a fanless design in a 132 x 91.5 x 68.5 mm format. They deliver up to 100 TOPS and support two PoE cameras for image capture and edge inference.

The slimmer AIB-MO23/33-S1 and AIB-MN33/43-S1 measure 127 x 85.75 x 28.45 mm. They deliver up to 157 TOPS and support dual MIPI or PoE camera configurations.

Across the series, the systems support wired and wireless connectivity through M.2 expansion, including LTE, 4G/5G, Wi-Fi and Bluetooth. They also include NVMe storage and a range of input and output options for direct connection to field equipment.

Aetina is positioning the systems for applications such as visual inspection, security monitoring and other edge inference tasks that rely on processing data near the source rather than sending it to centralised infrastructure.

Industrial settings

The Mini Series is intended for installations where larger computing systems are difficult to place. The systems accept 12-24V DC power and are rated for operation between -25C and +55C.

The fanless design is intended to reduce maintenance demands in industrial settings, where dust, vibration and restricted access can make active cooling less practical.

The launch reflects a broader push by hardware suppliers to package AI processing into smaller industrial systems as businesses look to run computer vision and multimodal models on site. Running models locally can help users manage latency, network dependency and data handling in factories, transport hubs and remote facilities.

The systems are compatible with NVIDIA JetPack 6.2 and later versions. Aetina also offers board support package customisation services and technical consulting for customers integrating the devices into their own systems.

Deployment focus

All four products are available for mass production, making this less a development-stage announcement than a commercial push aimed at customers preparing for volume deployment.

Industrial users have increasingly sought smaller systems that can handle both traditional vision workloads and newer multimodal AI models. In this market, support for camera inputs, industrial power ranges and compact enclosures can be as important as raw processing performance.

One intended use is lightweight vision-language model work, in which video inputs are combined with natural language processing for tasks such as recognition, prompt understanding and anomaly detection. That places the new products in a growing segment of the edge AI market that extends beyond image classification to systems that can interpret scenes and respond to instructions.

"AI deployment is no longer restricted to large-scale infrastructure. In actual field scenarios, high-performance and reliable intelligent computing are required within a limited footprint," said Troy Lin, Director of Edge AI Computing Product Division at Aetina.

"Through the Mini Series, Aetina provides a scalable edge AI platform that assists customers in rapidly deploying Vision AI applications from the development stage to actual production environments," said Lin.