Hosted by IDTechEx
Artificial Intelligence Research
An expert outlook on the world of AI
HomeEventsReportsTVCareersAbout UsSign-up or LoginIDTechExTwitterFacebookLinkedInYoutubeRSSForward To Friend
Artificial Intelligence Research
Posted on September 19, 2018

Deep learning computer vision solution for trains

poster ad small
Klas Telecom has announced a groundbreaking solution for the railway industry that delivers deep learning computer vision and object detection technology to the network edge and onboard trains. This new, multiplatform solution will allow railway network operators to deploy applications, such as video surveillance and smart cameras, that mimic human vision and perform tasks, such as vehicle identification, intruder detection and empty seat recognition, that result in improved safety, security and risk analysis. The new solution consists of the Klas Telecom TRX Connected Transportation Platform R6, a router/server and six-modem cellular gateway device, running Intel® OpenVINO™, enabled by KlasOS, convolutional neural networks (CNN) that together enable deep learning inference on the edge. Klas Telecom's TRX R6 is unique in the railway communications systems market in that it deploys with accelerated performance compute-intensive workloads, like the Intel® CNN-based machine learning models, in a single 4.5 kg, 250 mm x 279 mm x 76.5 mm chassis. TRX R6 is built to meet environmental compliance standards for rail, contains Intel® Core™ i7, i5 and i3 processors and provides an industry-leading 32 GB of RAM and up to 8 TB of built-in storage.
 
Klas Telecom is an Intel® IoT Solution Alliance member that specializes in embedding and extending enterprise-grade Intel® processing to austere edge environments. Customers interested in increasing deep learning performance can recognize the following benefits from deploying OpenVINO™ on Klas Telecom hardware:
  • Enabling CNN-based deep learning inference on the edge
  • Supporting heterogeneous execution across OpenVINO™ accelerators - CPU, GPU, Intel® Movidius™ Neural Compute Stick and FPGA - using a common API
  • Shortening time to market with a library of functions and pre-optimized kernels
  • Hardware acceleration
 
"It is exciting to be able to bring artificial intelligence and machine learning into railway use cases such as traffic monitoring, vehicle identification, pedestrian identification, real-time risk analysis, security and passenger information services," said Klas Telecom Director of Business Development for Transportation Brendan Fleming. "It means our customers can save time by using computers to perform intelligent functions and predictive analysis. Ultimately, time means money and we hope to save our customers both."
 
Webinars Generic Banner
Source: Klas Telecom
Top image: Wilson Times