What VR Needs to Get Right…VR and AR continue to struggle in the market. While CES 2020 demonstrated entrepreneurs and large companies still seeking a broader market, it also clearly reflected pull back from major players. Example: Samsung once sponsored giant VR experiences in Central Hall, but this year delivered only silence around their retired GearVR experiment.
Technology can always argue cart before horses, chickens before eggs. VR generates that dialog as well. Will the app drive the market or will better tech enable more willingness to adopt and develop, driving the market? The answer lies in the future, in that either will work. It depends on which one manifests first. Technology analysts will create a narrative in the future that reflects what did happen. The speculation will fall away. Until that happens, VR and, to a lesser extent, AR, sit like Schrödinger’s Cat in a state of intertwined probability waiting for the future to make a measurement so it can collapse into a representative reality.
In the meantime, we have compiled a list of what VR needs to get right still, with a bit more exposition on the chicken and the egg problem near the end of the post.
Six-Degrees-of-freedom (and movement through space). If you must sit to experience VR with a headset, then that headset does not support Six-Degrees-of-Freedom or SDoF. SDoF is what you experience in the real world: up, down, left, right, forward, and back. When in VR, people naturally want to lean into a scene or walk through it. That is why “VR” video creates such a subpar experience. Beyond SDoF is the general movement through large spaces. Teleportation and expensive peripherals like virtual reality-enabled treadmills bring some sense of movement, but neither proves ideal. The disconnect between the mental and physical experience remains a significant hurdle for VR developers and theorists. Geek.com’s Daniel Starkey (The big problem with VR no one is talking about) offers a perspective on what developers miss about movement in games. If they can’t get it right in games, they won’t get it right in other applications.
Controllers do not permit natural actions within virtual worlds. They may work for games as substitutes for guns or devices, but they work less well for building experience in enterprise solutions that require kinesthetic memory development.
Interpupillary distance (IPD). The distance measured in millimeters between the centers of the pupils of the eyes. Tired eyes and headaches ensure if vendors don’t get this right. Most headsets don’t provide hardware changes that shift the distance between lenses, and software fixes don’t cover all myopia and astigmatism issues.
Resolution. The availability of 8K resolution experiences on devices like those from XTAL makes a case for good VR experiences that deliver equal measures of fidelity and field of view. Higher resolution also helps with the suspension of belief and the elimination of the Screen Door Effect in most circumstances.
Field of view. The wider the field of view (FoV), the more immersive the VR experience. Smaller displays reduce FoV but may offer higher resolution. Ultimately FoV and Resolution should not be a trade-off.
Foveated rendering. This means that the software renders the center of the participants viewing area with higher fidelity than content relegated to peripheral vision. Foveated rendering requires eye tracking, fast processing, and robust algorithms capable of keeping up with the participant’s activity and content compiled to leverage the technology.
VR UI. Even though PC software devolves beyond the operating system into various interpretations of the core platform’s user interface rules, generally accepted norms make it reasonably easy to use the basic features of every program on an Apple or Windows device. Games vary much more widely than business apps, and because much commercial VR software ships as a game, or from game developers, each offers a unique view of the UI. Both applications require common ways to represent bodies, hands, and physical implements in the environment. Beyond the user in the space, VR requires a virtual entry point that makes participants feel comfortable navigating through the systems and a healthy batch of applications. Without an intuitive UI, people will feel overwhelmed when learning apps. They will either abandon the platform or adopt only a single app that performs a unique function without worrying about how to better leverage the investment.
Participant health impact. Health-related issues remain an open question until more of the items above get worked out. Some of the solutions may alleviate some health problems, but new technologies may also introduce new stresses. Current common issues facing VR participants include headaches, eye strain, nausea, and dizziness.
The personal relationship to hardware. Personally getting into and using SDoF VR remains convoluted. Add to that glasses that don’t fit inside a headset or fog up if they do, straps, and other accouterments that either cause physical irritation or pain or don’t fit well, leaving a headset, for instance, too loose for a pleasant experience. And then there is the tether to the PC for better experiences than any mobile solution. VR remains CPU and GPU intensive, so several elements above, like high resolution, foveated rendering and SDoF requires a PC to ignite the experience.
Other factors that cause more chicken and egg deliberation include:
Cost. We see price as a red herring. If VR solved the issues above and was still expensive, it would find a market. Many historically successful technologies started off costly, but that didn’t stop early adopters from creating solutions. Early applications in the PC market, like VisiCalc or Lotus 1-2-3, succeeded despite severe hardware limitations and high prices for hardware. Keep in mind that a fully loaded IBM PC ran about $6K. That is about $17K in current dollars. Even the entry-level IBM PC at around $1,600 would cost over $4,500 in 2020 dollars. People were willing to shell that out for DOS-based spreadsheets and word processing. If the VR world did really come up with a killer app, it might well drive the market despite hardware, software or experience limitations. PCs, however, didn’t introduce convoluted physical requirements or cause an immediate impact on health (carpal tunnel took a while to recognize) so the barriers to entry beyond price were less overwhelming.
The current pricing perception revolves around VR price points that don’t deliver perceived value equal to the price (and effort) associated with the hardware and software purchase.
Bandwidth for streaming. Unless all content arrives from a local data store, at some point VR will require streaming data, and streaming requires significant bandwidth. Streaming will require faster in-home and business connections as well as compression approaches. As seen with television, because display technology continues to improve, at some point content will be asked to keep up with capability.
The killer app problem. Many analysts assert the slow adoption of VR relates to the lack of a killer app. Any killer app today, created under the constraints above, will result in subpar application experience. So, we don’t see a killer app on the horizon that will entice people to buy into VR. Improved experiences and technology will also encourage more extensive experimentation, which may lead to a broader range of use cases. Most perceive killer apps as emergent. An app that only works because VR works. If there is another, perhaps better way to do something, then that approach will likely win any complementary, but not compelling VR version. Training for working safely in physical spaces, like factories, looks compelling on the surface when the cost of structure outweighs experience, but VR offers little that cannot be done better in a real physical simulation where hands, heads, bodies, and gravity all interact the way they would in the field.
Back to the data problem. Serious Insights believes that killer apps already exist, but that the data architecture of the world does not yet support their implementation. VR ideally draws from data that represents the real world, cities, buildings, and machines. Most models today are superficial. Even complex models often model only one aspect of a thing, like the exterior of a building, not its pipes, wiring, ducts, etc. Those things exist, but not in an integrated model that would allow emergency response or building maintenance to leverage them.
VR also needs to define its privacy model.
Personally identifiable information (PII)isn’t VR or AR specific but needs addressing by the VR vendors. What people look at in virtual spaces will prove vital to advertisers and other analysts of in-game behavior. What the PII data is, how consent gets captured, and what people do with the data will shape perceptions of VR and AR’s safe spaces profile.
A note on augmented reality (AR)
AR presents its own issues including:
Occlusion-recognition tracking. How the software handles physical things that exist in the AR field of view.
Power. AR is highly energy-intensive. Be it a phone or a headset, constant AR use drains batteries quickly.
Device-based solutions. If AR ends up primarily being delivered on smartphones and tablets, several hands-free use cases get lost in the mix. AR’s best implementations allow people to react to the information with their full body.
For more on virtual and augmented reality from Serious Insights click here.
Daniel W. Rasmus
Daniel W. Rasmus, Founder and Principal Analyst of Serious Insights, is an internationally recognized speaker on the future of work and education. He is the author of several books, including Listening to the Future and Management by Design.