Virtual reality (VR) is a sister technology of AR. As described, AR augments your current experience in the real world by adding digital data to it. In contrast, VR magically, yet convincingly, transports you to a different (computer-generated) world. VR is intended to be a totally immersive experience in which you are no longer in the current environment. The sense of presence and immersion are critical for VR's success.
AR does not carry that burden of creating an entire world. For AR, it is sufficient for computer-generated graphics to be added to your existing world space. Although, as we'll see, that is not an easy accomplishment either and in some ways is much more difficult than VR. They have much in common, but AR and VR have contrasting technical challenges, market opportunities, and useful applications.
Since VR is so immersive, its applications are inherently limited. As a user, the decision to put on a VR headset and enter into a VR experience is, well, a commitment. Seriously! You are deciding to move yourself from where you are now and to a different place.
AR, however, brings virtual stuff to you. You physically stay where you are and augment that reality. This is a safer, less engaging, and more subtle transaction. It carries a lower barrier for market adoption and user acceptance.
VR headsets visually block off the real world. This is very intentional. No external light should seep into the view. In VR, everything you see is designed and produced by the application developer to create the VR experience. The technology design and development implications of this requirement are immense. A fundamental problem with VR is motion to photon latency. When you move your head, the VR image must update quickly, within 11 milliseconds for 90 frames per second, or you risk experiencing motion sickness. There are multiple theories why this happens (see https://en.wikipedia.org/wiki/Virtual_reality_sickness).
In AR, latency is much less of a problem because most of the visual field is the real world, either a video or optical see-through. You're less likely to experience vertigo when most of what you see is real world. Generally, there's a lot less graphics to render and less physics to calculate in each AR frame.
VR also imposes huge demands on your device's CPU and GPU processors to generate the 3D view for both left and right eyes. VR generates graphics for the entire scene as well as physics, animations, audio, and other processing requirements. Not as much rendering power is required by AR.
On the other hand, AR has an extra burden not borne by VR. AR must register its graphics with the real world. This can be quite complicated, computationally. When based on video processing, AR must engage image processing pattern recognition in real time to find and follow the target markers. More complex devices use depth sensors to build and track a scanned model of your physical space in real time (Simultaneous Localization and Mapping, or SLAM). As we'll see, there are a number of ways AR applications manage this complexity, using simple target shapes or clever image recognition and matching algorithms with predefined natural images. Should this be: Custom depth sensing hardware and semiconductors are used to calculate a 3D mesh of the user's environment in real time, along with geolocation sensors. This, in turn, is used to register the position and orientation computer graphics superimposed on the real-world visuals.
VR headsets ordinarily include headphones that, like the visual display, preferably block outside sounds in the real world so you can be fully immersed in the virtual one using spatial audio. In contrast, AR headsets provide open-back headphones or small speakers (instead of headphones) that allow the mix of real-world sounds with the spatial audio coming from the virtual scene.
Because of these inherent differences between AR and VR, the applications of these technologies can be quite different. In our opinion, a lot of applications presently being explored for VR will eventually find their home in AR instead. Even in cases where it's ambiguous whether the application could either augment the real world versus transport the user to a virtual space, the advantage of AR not isolating you from the real world will be key to the acceptance of these applications. Gaming will be prevalent with both AR and VR, albeit the games will be different. Cinematic storytelling and experiences that require immersive presence will continue to thrive in VR. But all other applications of 3D computer simulations may find their home in the AR market.
For developers, a key difference between VR and AR, especially when considering head-mounted wearable devices, is that VR is presently available in the form of consumer devices, such as Oculus Rift, HTC Vive, PlayStation VR, and Google Daydream, with millions of devices already in consumers' hands. Wearable AR devices are still in Beta release and quite expensive. That makes VR business opportunities more realistic and measurable. As a result, AR is largely confined to handheld (phone or tablet-based) apps for consumers, or if you delve into wearables, it's an internal corporate project, experimental project, or speculative product R&D.
