Demonstrating edge video processing for tactical environments
Government organizations rely on full-motion video (FMV) to support intelligence gathering around the world and around the clock. Unlike commercial media and entertainment markets, where content is tuned to perfection through stages of preproduction, production, and postproduction, government organizations often generate and consume content on the fly, in near-real time. If the tempo alone wasn’t challenging enough, government video can originate in some of the most remote areas of the world, in extreme environments, and is often time sensitive and/or mission critical. Deploying effective video processing techniques in a live FMV architecture not only supports high image quality and reliable performance, but improves accessibility when it matters most.
As aerial imaging systems extend situational awareness from above, disseminating this content to localized and enterprise users can be a challenge. AWS Elemental is addressing these complex environments by leveraging commercial cloud technologies in a hybrid, rapidly deployable AWS cloud architecture that supports connected and disconnected continuity of operations.
AWS Elemental has teamed up with several government organizations and agencies at the 2019 USD(I)-sponsored Enterprise Challenge exercise to demonstrate several key uses cases for Tactical Edge Video Processing in mission focused scenarios. The event will serve as an operational proving grounds, applying cloud-enabled technologies against real-time simulated mission environments:
- When a natural disaster strikes, crippling ground-based infrastructure, how will emergency responders be augmented with video from above to support their mission?
- In tactical, forward-deployed military environments, how can aerial footage be furnished to troops on the ground as part of an overwatch mission?
- In a search-and-rescue effort at sea, how will airborne-equipped sensor systems provide another set of eyes to vessels on the move?
AWS Elemental will demonstrate various aspects of a hybrid cloud workflow leveraging AWS services aimed at improving the FMV quality of experience over constrained ground-based RF networks, distributing reliable video streams to local tactical users, and performing protected contribution of live FMV content to enterprise stakeholders via AWS Cloud Regions. The architecture includes a number of AWS services in use, including AWS Elemental Live on-premises encoding systems, combined with AWS Elemental MediaConnect, AWS Elemental MediaLive, and AWS Elemental MediaPackage for cloud contribution, packaging, and delivery. In addition, nearly a half-dozen APN Partners will be involved in the exercise, including ISR (intelligence, surveillance and reconnaissance) sensor systems utilizing AWS Elemental Live Small Form Factor HEVC/H.265 encoding on platform, and PED (processing, exploitation, dissemination) video/geospatial applications running on Amazon Machine Images powered by AWS Snowball Edge Compute and Storage Optimized devices.
In upcoming blog posts, we’ll walk through different aspects of this proof of concept architecture, addressing specific technical considerations in the following areas:
- Intelligence, Surveillance, and Reconnaissance video contribution and distribution: from Air-to-Ground, Ground-to-Cloud, & Air-to-Cloud;
- Best practices in selection of video formats, resolutions, compression techniques, etc.
- Selecting the right network transmission protocols (e.g. Adaptive Bit Rate vs UDP)
- Streaming tactics for Disconnected, Intermittent, and Limited-bandwidth (DIL) environments
- Tips and tricks utilizing PED-centric applications running on AWS Snowball Edge storage and compute devices.