Since Adlink’s collaboration with Nvidia as a Jetson Preferred Partner, the company has launched several specialised controllers and systems that feature high processing power for AI applications and machine learning. These include the EOS-J system, which has an integrated Jetson TX2 chip and four GigE or USB 3.0 interface options for connecting cameras or frame grabbers.
Edge computing can often be a more efficient architecture in industrial environments. Taking AI decisions directly at the machine, the robot or the autonomous vehicle allows for quicker actioning than forming data communication with the cloud. Development trends such as autonomous driving and robotics applications show this in the context of vision computing. In both these cases, “seeing” and “recognising” have the greatest impact on fast on-site decisions. Machine learning and deep learning optimise quality control with visual inspection, classified errors and enable predictions and trend developments through machine diagnosis.
Acceed has added two models of the new EOS-J vision systems now. The two models only differ by the offering either four type GigE or USB 3.0 camera ports. The EOS-J is a compact system together with the cameras or frame grabbers and it provides 32 isolated IO interfaces for use in edge computing.
In addition to the Jetson chip, EOD-J also offers a Quadro GPU with Nvidia Pascal architecture. The integrated Cortex-A57 processors from ARM 256 CUDA cores support robust deep learning and inference applications.
Application-specific image processing solutions can be developed easily with neuronal models, optimised with Nvidia’s Digits or TensorRT tools. The developments can be directly applied on EOS-J. A gigabit Ethernet connection and two communication ports as well as an HDMI output support integration in the field.