How AI-Based Automation Enables VR and AR Experiences

How AI-Based Automation Enables VR and AR Experiences
Azz-Eddine Mansouri, General Manager at Ciena Middle East
Published: 15 May 2019 - 9:35 a.m.

We are turning the page to a new chapter - the fourth industrial revolution - thanks to rapid advances in technology, including Augmented Reality (AR) and Virtual Reality (VR).

According to IDC, the worldwide shipments of augmented reality and virtual reality (AR/VR) headsets are forecasted to reach 8.9 million units in 2019. In addition, by 2023 virtual reality headsets are expected to reach 36.7 million units. On the other hand, the AR headset market is poised to reach 31.9 million units in 2023. Indeed, both emerging technologies have quickly gone mainstream in recent years and will only make more headway into both the enterprise and consumer worlds across healthcare, manufacturing, entertainment and other industries. In fact, according to IDC, AR and VR technologies are expected to grow $6bn by 2020 in the Middle East and African market. The UAE, unsurprisingly, is at the forefront in implementing advanced solutions for its industrial sector.

Demystifying Virtual Reality

VR is a 100 percent virtual, simulated experience. This emerging technology is exceptionally sensitive to lag and slowdown—delays between when an input is placed and when the system reacts to it, and noticeable disruptions in the consistent stream of data being delivered, respectively. A significant portion of its value proposition includes the experience of actually being transported somewhere, and thus a frozen screen or patch of pixelated haze smashes that illusion quickly, ruining the experience for many—and in some cases causing motion sickness.

When an event is broadcast in VR— for instance the Expo 2020 Dubai, which will offer innovative digital experiences to attendees—camera rigs that capture 360-degree (or 180-degree depending on the event) panoramic views are needed to provide the viewer with the ability to look at every angle. This requires a number of lenses and thus multiple video streams moving side-by-side. To transmit this information, copious amounts of bandwidth are required—up to 4 to 5 times as much for 360-degree video compared to regular video, according to YouTube’s Anjali Wheeler.

Furthermore, how the content is being streamed as ‘live’ or not also impacts the bandwidth requirements which will be significantly higher in this case.

It is also important to differentiate the two types of ‘live’ VR – ‘live’ as in watching an event as it occurs; or ‘live’ involving interaction with others within a virtual environment. In the first format, the unit is passively accepting the data stream from the network, which requires low latency and a high bandwidth connection to achieve high video throughput. With the latter, to enable interaction between the VR source and multiple users, low latency is required so there is no noticeable delay, even as data is moved back and forth between the individual VR units connected and servers.

Augmented Reality: Blurred Lines between Real and Digital

The crucial difference between VR and AR lies in the way digital content is mixed with reality. Unlike VR, AR allows users to integrate elements of reality into their experience. For example, AR can work through mobile phones using the integrated camera and is the kind of virtual reality that powers Pokémon Go. The game’s massive adoption showed the average consumer’s appetite for AR technologies.

There needs to be a way to deliver AR content to devices with high bandwidth and low latency while allowing compute to happen outside the unit. According to GSMA Intelligence, generic AR applications will require upwards of 100Mbps bandwidth throughput and nearly as low as 1ms delay—difficult specs for a device you can walk around with.2

As the increase in demand for network bandwidth continues, service providers must bring new approaches to how they design, deploy, and operate their networks so they don’t compromise a great customer experience. Technologies like Software-Defined Networking (SDN) and Network Functions Virtualization (NFV) are designed to help deal with some of these pressures.

So will machine learning, a form of AI that uses statistical techniques to enable systems to ‘self-learn’ and progressively improve network performance on a specific task. For example, Blue Planet’s Proactive Network Operations solution uses advanced machine learning algorithms to pinpoint potential issues before they occur and instructs the network the best course of action to take to resolve.

Enabling emerging the digital experiences through AI-Based Automation

The success of AR and VR doesn’t solely depend on hardware; it requires intelligent software capabilities. As more companies move to digital, they adopt software-based approach which relies on AI. In fact, according to Gartner, the adoption of AI in organizations has tripled in the past year.

The adoption of AI is a great enable for several businesses. For instance, Over-the-top (OTT) content and service providers such as Google, Facebook, Netflix, and Amazon are leveraging big data analytics and AI-based automation in their services to provide every customer, a highly customized experience.

Thus, by leveraging AI and an intelligent automation platform, service providers can dynamically allocate resources to specific demanding applications, such as those run for AR/VR. In addition, intelligent automation, built on software-defined control and orchestration with integrated analytics—and one that should be on every provider’s road map. Seamless AR and VR experiences demand the highest bandwidth possible and with AI, applications for AR and VR devices can be configured across the network so they receive the service they require.

As AR/VR continue to find use cases in our day to day lives end-customers have increasingly high expectations for their quality of experience. Therefore, enterprises need to ensure technical bedrocks such as AI are in place to be able to deliver a satisfying user experience.

Click here to add your comment

Please add your comment below
Your email address will not be published