Rugged Inference at the Edge
October 23, 2019
Concurrent Technologies announces their involvement and progression in the development of Artificial Intelligence (AI) compatible rugged solutions.
These solutions are designed to provide Inference at the Edge capabilities in the type of uncontrolled and challenging environmental conditions typically found in intelligence gathering equipment within the defense, transportation and exploration markets. These types of devices use a multitude of sensors and, over time, the resolution and accuracy of these sensors has vastly improved multiplying the amount of data dramatically. Historically the data gathered by this type of equipment was analyzed after the event in a prolonged and manual way, leading to delays and unreliable decisions. Inference at the Edge provides a way to analyze the data gathered by the device in the field to make educated and reliable decisions in real time.
Concurrent Technologies has now tested some of its recent rugged processor boards like TR H4x/3sd-RCx for use as inference engines. This process was done using trained neural networks from a variety of models such as TensorFlow™, MXNet™ and Caffe to analyze the data accurately, quickly and efficiently. The processor boards are tested compliant with the Intel® OpenVINO™ toolkit. This includes a model optimizer tool to facilitate the transition between the training and deployment environment, performs static model analysis, and adjusts the models for optimal execution on the Concurrent Technologies boards.
Jane Annear, Managing Director, commented: "It's an exciting time as we continue our journey to provide more solution capabilities based on our existing products. Artificial Intelligence is expected to be an effective way of concluding answers to the difficult scenarios that face many of our customers as they analyze data in real time."