How Edge Computing Unleashes Innovation in Live Streaming

(Image credit: Videon)

Live video streaming is taking off with the global video streaming market estimated to exceed $240 billion by 2030. Driven by the media and entertainment industries which are joined by an ever-increasing number of vertical use cases, it’s hard to think of a sector where live video streaming isn’t a fundamental parameter. 

However, video itself is changing. Once a passive medium delivered to be “consumed” by audiences, video today is an actionable resource that facilitates processes, enables analytical insights, and, in many cases, generates new income.

But as live video streaming is changing rapidly, transforming industries, and opening up new opportunities to content owners, the ways in which media companies capture, process, and distribute live video streams needs to be reconsidered. For video to become actionable, traceable, and monetizable, media industry players need more control and more freedom to shape the workflows they actually require and distribute video to all the right destinations.

Edge computing at the point of video origin is a paradigm shift, taking live video to the next level.  

The problem with live video workflows
Historically, media organizations have relied on traditional broadcast-grade equipment capturing and encoding video. While reliable, these devices are also costly and largely inflexible as they are designed for TV applications and not smaller-scale events. 

On the other hand, there are the lower-end video encoders utilized by content creators for platforms like YouTube, TikTok, and Twitch. However, these encoders are tailored for specific use cases and only offer limited reliability, functionality, and flexibility. 

In nearly every streaming video application, the cloud is relied upon for a significant part of the solution, be it for higher-end broadcast TV workflows all the way down to low-end workflows.  The cloud overcomes compromises of the video encoder, acting as an intermediary for processing video streams and distributing them to the right platforms. However, these workflows are still subject to delays as they add extra steps to the video supply chain. 

For some use cases, like sports betting, latency can be a real deal-breaker. Not to mention that video compression before sending to the cloud can damage quality when the cloud is needed for a second level of video decoding and re-encoding. 

But issues with speed and video quality are accompanied by the challenge of cost. The cost of cloud-based workflows can also prove prohibitive for media industry players that need to scale their cloud processing to several hours a day over a long period of time. 

In addition, cloud workflows rely on IP connectivity, either broadband bandwidth or 4G/5G networks. Moving raw video to the cloud for processing and often sending it back to the source requires connectivity to be scaled to ensure the video quality — this is a cost-intensive process.

Changing the dynamics of live video with edge computing
Edge computing at the video source revolutionizes live video by delivering unprecedented levels of flexibility, reliability, scalability, high quality, and low latency. 

All this by bringing powerful processing power to the point where video is created. Instead of media companies deploying a fixed device that performs one video processing task, they can leverage an edge computing-powered, programmable device that can perform any task based on its available processing power, storage, and software applications. If the user wants to change a workflow, they just need to adapt the software without replacing the device. 

This means that media companies gain control and freedom over their workflows adapting them to their actual business needs instead of the other way round. When considering the potential of cognitive technologies like Artificial Intelligence (AI) and Machine Learning (ML), metadata, etc. the opportunities for innovation are endless — especially as placing the edge computing capability as close to the video source as possible, dramatically reduces latency and transit costs. 

In several use cases, edge computing for live video compliments the cloud by running many of the video processing tasks locally, augmenting cloud functionality. The video supply chain becomes more reliable, efficient, and better for content owners and audiences.

Edge computing for live video in action
The 2019 FIVB Volleyball Men’s World Cup was the 14th staging of the tournament, contested by the senior men’s national teams of the members of the Fédération Internationale de Volleyball (FIVB), the sport’s global governing body. The tournament was held in Japan in October 2019 and televised by several broadcasters and via streaming services such as Fuji TV.           

The streaming service identified a number of needs, including low-latency streaming between 2 seconds and 6 seconds, similar to broadcast delay. However, customers were not willing to pay more for low latency than they currently pay for traditional streaming. Fuji TV also requested standards-based technologies with a variety of player, encoder, and delivery options as opposed to proprietary solutions. 

The deployment needed to also offer synchronized delivery across viewers and platforms via traditional CDNs. The solution, designed to support tens of thousands of simultaneous viewers, used a combination of Videon EdgeCaster and its edge processing capability to provide low-latency encoding in a low-power, small form factor footprint to deliver DASH CMAF packaging to AWS Elemental MediaStore.    

Video flowed onward to streaming sites and apps via Amazon CloudFront distribution using the AWS backbone for low-latency, low-packet loss, high-availability transmission from origin to edge. This was combined with the VisualOn Player to provide a stable video experience for viewers offering synchronized viewing across broadcast and OTT platforms. 

The whole solution was built and deployed in just three weeks, with an average glass-to-glass latency of 3 seconds, and provided stable distribution over the two weeks of the World Cup. 

Welcome to the world of endless innovation  
Edge computing at the point of video origin marks the beginning of a new era for live video streaming. Placing raw processing power where the video is created and empowering users to build the functionality they need when they need it at a very low cost is a real game-changer. 

For the first time, industry players have true control and freedom over their workflows without being limited by latency, scalability, or cost. With live video streaming booming, media companies and content owners can leverage edge computing to unlock unlimited opportunities.