Aetina Brings Compact Edge AI Computing to Vehicles, AMRs and Industrial Automation

Computing power is no longer considered a concern  as AI-powered vision is shifted from the cloud to real-world machines. Whats increasingly becoming a must is the ability to have this form factors compact enough to fit in tight applications, process multiple camera feeds simultaneously, and perform consistently in demanding environments. To address that challenge, Aetina is addressing it with its new AIE-VN34/44 and AIE-VO24/34 in-vehicle edge AI systems.

The palm-sized platforms feature NVIDIA Jetson Orin NX or Orin Nano modules with up to four GMSL2 cameras. The higher-performance AIE-VN34/44 configuration can achieve up to 100 TOPS, and the AIE-VO24/34 can achieve up to 67 TOPS. The systems measure only 136.3 x 132 x 63mm, making them an ideal fit for applications where standard industrial computers might not fit. 

In perception-intensive applications, the four GMSL2 camera interfaces stand out. 

They allow low latency, long-distance camera connections, and provide the capability to process multiple video streams locally instead of sending raw video to the remote server. It provides opportunities for applications like surround view monitoring for vehicles, blind-spot detection, obstacle avoidance and real-time situational awareness. The same architecture can go beyond the vehicle for industrial operators. Aetina sees applications for autonomous mobile robots (AMRs), smart factories and automated production lines, where compact edge computing can be used to support machine vision without taking up a big chunk of computing space. 

The hardware has been designed to handle field conditions as well. The shock and vibration testing is MIL-STD-810H, and input power range is 9-36V DC wide. A fanless design allows for operation within a -25C to 55C range and the ignition power control is designed to assist with safer power up/shut down in mobile applications.