News & Events

The role of gigabit networks in Virtual Reality

Virtual Reality launched to great fanfare in 2016, but reports of lackluster sales show it falling short of most industry projections. Slower adoption owes more to a lack of experience in developing content for VR platforms and the limits of current hardware or networks than to a real lack of interest in VR itself. High end headset prices and difficult setups limit these devices to enthusiasts and early adopters. Current top of the line hardware just passes the threshold of minimum recommended specs for VR. Entry level headsets like the Gear VR or Daydream are limited by the variations and capabilities of the mobile devices that drive them. A critical, but seldom discussed, missing component of a the ecosystem needed for VR to thrive is availability of advanced gigabit networks.

Once the difficulties of hardware and pricing resolve themselves through normal market evolution, consumers are still going to be limited to the largely canned experiences of installed software. Achieving dynamic and interactive experiences in the VR space is going to require quickly moving vast amounts of data and media to provide “as in life” real time experiences between users, all of which are unique features of advanced gigabit networks.

One of the best current examples of the limits of networked data in VR is demonstrated by the world modeling in Google Earth VR. This application brings the entire world to the user as if they are standing on an elaborate Earth-sized scaled model. Terrain and buildings are rendered by moving massive amounts of texture and 3D data corresponding to any point on the globe to a VR headset, but the experience is far from real-time. Standing in a new location, the environment first appears as giant blobs of blocky terrain and misshapen buildings that gradually sharpen and take form as better data streams in from the Google Earth data centers. The end result is undeniably amazing, but I liken the experience of standing there while a new area comes into focus with waiting for web pages to load over modem in the late nineties. Networks will need to be faster by an order of magnitude before we can realize the dream of flying over a beautifully rendered globe in real-time. Reliable gigabit networks are what is needed just to proper render static models and data. Looking forward to the integration of live data sources, like real-time traffic or power grid health, the need for high-speed and low-latency networks becomes even greater.

PlanIT Impact is an advanced architectural and civic planning application that demonstrates the potential of merging real time city data with VR environments. Currently, it is a web based tool that allows users to “better understand impacts to energy, water use, stormwater and transportation, balanced with economic impacts and Return on Investment, to enable Data-Driven Green Design.” It already requires gigabit networks on the back end to bring in the big datasets needed for detailed simulations. Plans to develop a VR based experience for more immersive planning environments will require even faster networks to support adaptive analytics based on users’ changing 3D environments to visualize the impact of different planning options. The accessibility and dynamic nature of this project promises to deliver unprecedented levels of planning efficiency, positive environmental integration and high levels of citizen engagement. Bringing in big data for dynamic analysis, merging architectural data about a city environment, and pushing that to a VR rendering platform is not a task for current networks, and PlanIT Impact is just one of an emerging ecosystem of applications that require the development of gigabit networks to be fully realized.

Current networking also limits current VR experiences to single user activities and will require ultra-low level latency to achieve real-time shared experiences for multiple users. An early example of this kind of multi-user interactivity is the VR Solar Energy education platform presented in a previous blog post. In this example, the 3D avatar of an instructor is rendered in real-time from a classroom and projected into a VR environment to engage with students as they interact with the learning environment. Another example of the potential of interactive VR apps is the ViatoR VR Language learning app; we demoed a prototype at the 2016 US Ignite App Summit in Austin. Here, interactivity with objects in real time and immersive language audio environments can be enhanced significantly by placing native speakers and learners together in the VR space. Other applications are starting to be developed that allow users to collaboratively sculpt a 3D object together or play competitive virtual sports like ping pong. Facebook demonstrated an excellent example of the potential of low-latency interactions in VR space at Occulus Connect 2016. In a live demonstration, three people interacted in real time handing each other virtual objects and even fencing with one another in a shared VR environment. These demonstrations look great on stage, but before they can become a reality for the majority of users, networks will need to evolve to provide sub ten millisecond responsiveness without interruptions or jitter, so physical actions by the participants merge seamlessly with objects in the virtual environment.

Augmented Reality (AR) and VR are often referenced together and AR can have a much higher requirement for low-latency than VR. In AR, you use the natural environment and overlay data or render objects into the user’s view of the real world, in real-time. This is fine for simple rendering of static objects, but the moment you need AR objects to be interactive with changing objects in the real world, the timing of that interaction has a hard real-time requirement. An early example of the potential of mixing real-time data in AR was demonstrated at the fall 2016 KCVR Hackathon in Kansas City by a prototype called Data X-Ray. The application merges real-time data analytics with objects in the world to give an accurate picture of the state of a network device as they are happening. The power of AR to overlay data on real world objects as they carry out operations is apparent, and implied in that is the critical need for the kind of low-latency that links data points to the moment a real world object changes as a result. Extrapolating this concept to a car manufacturer getting data about a vehicle as it drives by, or Public Safety officials getting data about environmental hazards as officers are interacting with the environment, the criticality of reliable low-latency networks becomes apparent.

Until hardware specifications catch up to the needs of emerging VR markets and content developers better realize the use of VR as a platform, discussions on the role of advanced gigabit networks is likely to remain muted. These issues are relatively easy to address as they are controllable at the level of individual users and developers. Network improvements will require a broader national approach and thoughtful implementation to advance beyond the variations of the page load we have on today’s internet to a dynamic and interactive environment teased with the advent of true VR technologies.