Edge Video Analytics reveals the intelligence of networks of the future

As part of the Edge Video Analytics experiment, carried out during the 2018 Roland-Garros French Open, Orange is making its networks take on a new role, favouring dynamic processing of information rather than transporting it in order to provide a subsequent analysis. Explanations.

“Edge Video Analytics anticipates the power of the networks of tomorrow, as it heralds their ‘softwarisation’ and distribution”

Today, we transport information. With this Edge Video Analytics experiment, we are converting this information as close to the source as possible, so that only useful information is transported,” summarises Jean-Pierre Casara, Director of Network Innovations of the Future within Orange.

Making the images talk in real time

Specifically, the experiment consists of installing cameras connected to the Suzanne Lenglen court, every one of which points to a section of the stadium (“tribune”). The images captured by these cameras are fed back to the mobile network and are interpreted in a decentralised mini-Cloud located in the same stadium. Here, an artificial intelligence solution converts these images into the indicator required, namely, the precise number of spectators present in every area of each “tribune”. These indicators, around 30 in total, are updated every minute and fed back via a website to the staff of the Fédération Française de Tennis (French Tennis Federation – FFT). Combined with other information, such as the progress of the current match, it therefore allows  the flows of spectators so that it is possible for them to dynamically adapt reception, orientation and seating management, etc. This will result in an optimised experience both from the spectator’s and organiser’s point of view.

5G and Edge Computing – the new paradigm

Today, the experiment is taking place with the help of a private 4G network while awaiting the arrival of 5G. “Edge Video Analytics anticipates the power of the networks of tomorrow, as it heralds their ‘softwarisation’ and distribution,” explains Jean-Pierre Casara. Indeed, the principle of  edge computing, which is at the heart of the solution, involves distributing intelligence and computing power to bring it as close to the source data as possible. Thanks to its ability to process locally and in real time, edge computing therefore has the advantage of limiting the data transported to the Cloud. As such, it never leaves the mobile network, which allows for increased commitment with regard to respecting privacy and data.

Jean-Pierre Casara specifies that market prospects are important, and operators have a role to play. “There is a real enthusiasm for processing images, their interpretations and predictive models. New markets can open up for developers of specialised applications in these fields and Orange could be there to provide the means by which these applications are operated within networks of the future.”

Read also on Hello Future

New algorithmic challenges for network virtualisation

Discover

Sylva: A Common Software Framework for Network Interoperability via the Cloud

Discover

Telco Software: A New Era for Carriers

Discover
Live streaming has become increasingly widespread. With the addition of 5G, this service can be dramatically improved at all levels, including image quality, download times, interruptions and lag. Faced with today’s generations’ enthusiasm for live feeds, researchers are now working to adapt live streaming TV so it can be done on the go. The Goal: Lag-Free Live Streams Getting closer to what’s happening live is one of the main challenges in the field of live streaming. Yet, streaming over the Internet using Wi-Fi or 4G still results in a lag of 30, 40 or even 50 seconds on tablets or smartphones. This lag will particularly hit home for any soccer fans who have ever heard their neighbor watching TV and cheering for a goal they haven’t seen yet. It also affects participants in time-limited interactive TV game shows and televised broadcasts by figures of authority in relation to announcements, alerts or disasters, for example. Ensuring service continuity, particularly when faced with high demand, is another challenge of live streaming. At Orange Innovation, researchers are therefore thinking about how they can make improvements in the field of TV streaming on the go, using a combination of 5G, video streaming technologies (multicast, low latency), network bandwidth allocation (network slicing) and edge computing. Their work has primarily focused on mutualizing streams; a key way of saving bandwidth. Dominique Thômé, Product Manager Innovation Data TV, explains that “Unlike unicast technology, which broadcasts streams as many times as there are simultaneous connections, multicast should allow a single stream to be broadcast to thousands of people connected to a large 5G zone. This mutualization prevents bandwidth loss and, consequently, service interruptions from network congestion. Another advantage, which is of great importance to Orange, is that it consumes less energy and therefore contributes to the transition to a low-carbon economy.” Recognizing the Know-How of Carriers Experiments carried out in the Orange laboratory have yielded interesting results. A real-time readjustment of video quality to prevent network saturation resulted in each customer being able to watch TV with only five seconds of lag, confirming the feasibility of 5G live streaming on the go. In fact, faced with ever-increasing volumes, some broadcasters are beginning to turn to carriers to broadcast their TV streams. They need players that are able to transmit this huge amount of data while ensuring optimal quality, in order to avoid any latency problems. Thibaut Mathieu, Director of Innovation for Interactive & Multiscreen Services at Orange says that “Our pioneering approach toward 5G live streaming highlights the valuable role that network carriers play, right at the heart of the system, compared to OTT players (“over the top,” such as the Tech Giants), both in terms of technology and business. We will be able to get involved in data transmission, with optimal mutualization technology that will save money and energy.” These technologies are consistent with Orange’s CSR commitment, both in terms of carbon footprint (lower energy consumption) and inclusion (broadcasting the right information at the right time). More than Just Entertainment The challenge goes far beyond the traditional TV broadcasting market itself. In the context of the health crisis, brands have been quick to understand the value of live streams to generate sales and are starting to venture into “Live Shopping.” Originating from China, this large-scale approach to teleshopping consists of an online event where presenters, influencers or personalities showcase products live to a digital audience who are able to order products or ask questions. Live Shopping is attracting more and more brands around the world. “With hundreds of thousands of people connected at the same time, its large scale will certainly create capacity issues” says Thômé. “This is another case where mutualization will ensure quality of service.”

How 5G Is Revolutionizing Live Streaming

Discover

A Lab for Testing the Benefits of Edge Computing on a 5G Network

Discover

MWC 2021: With Pikeo, 5G Networks Switch to the Cloud and Become Autonomous

Discover

Cloud computing all across the world

Discover

Vendée Globe: Innovative ways to immerse viewers in the heart of the action

Discover