Currently, corporations struggle to safeguard their workers from accidents and monitor remote operations such as oil fields, oil pipelines, transmission lines, transportation goods, machinery, plants, etc. for malfunctions. Often broken equipment, pipeline, or other nefarious activity goes unnoticed for a long period of time causing a great amount of loss. With computer vision and deep learning models on edge devices, it becomes easier and cheaper to surveil and correct anomalies in time to prevent losses using an end-to-end video analytics solution.
With this platform, camera(s) at the edge register the image, process it, and analyze it. If an anomaly is detected, they generate an alarm/message to take action. Otherwise, the information is sent to an on-premises computer, and then on to the cloud for further processing. This includes analyses such as tracing the subject and training the AI model. If the network is slow, offline, or power is down, then contingencies need to be made so that important information is not lost.
The challenge is to optimize and balance this end-to-end solution for multiple contingencies, including poor power and network conditions, number and type of cameras, number of intelligent gateways, on premise and off premise storage and platform requirements, balancing the AI workload (selection of right inference algorithm), total solution cost, and response time. All these challenges can be addressed by quickly modeling and simulating end-to-end solutions using Intel® CoFluent™ technology for the IoT.
Using Intel® CoFluent™ technology, users can quickly optimize their infrastructure configuration, resource utilization, cost, and performance.
Additionally, users can identify and remove deadlocks:
Overall, Intel® CoFluent™ technology with Intel® AI Products can be used to model and simulate an optimal video analytics and AI solution.
Notices and Disclaimers
Intel and Intel CoFluent are trademarks of Intel Corporation or its subsidiaries in the U.S. and/or other countries.