OSVR is an open source software platform and VR headset. Sensics and Razer launched OSVR 18 months ago with the intent of democratizing VR. Our goal is to provide an open alternative to walled-garden, single-device approaches.
Yuval is CEO of Sensics and co-founder of OSVR. Yuval and his team designed the OSVR software platform and built key parts of the OSVR offering. He frequently shares his views and knowledge on his blog.
It turns out that others share this vision. We saw exponential growth in participation in OSVR. Acer, NVIDIA, Valve, Ubisoft, Leap Motion and many others joined the ecosystem. The OSVR headset—called the Hacker Development Kit—has seen several major hardware improvements. The founding team and many other contributors have expanded the functionality of the OSVR software.
I’d like to describe how I hope to see OSVR develop given past and present industry trends.
Increased Device Diversity, More Choices for Customers
An avalanche of new virtual reality devices has arrived. We see headsets, motion trackers, haptics, eye trackers, motion chairs, and body suits. There is no slowdown in sight: many new devices will launch in the coming months. What is common to all these devices? They need software: game engine plugins, compatible content and software utilities. For device manufacturers, this software is not a core competency but ‘a necessary evil’. Without software, these new devices are almost useless.
At the same time, content providers realize it’s best not to limit the their content to one device. The VR market is too small for that. The more devices you support, the larger your addressable market becomes.
With such rapid innovation, what was the best VR system six months ago is anything but that today. The dream VR system might be a headset from one vendor, input devices from another, and tracking from a third. Wait another six months and you’ll want something else. Does everything need to come from the same vendor? Maybe not. The lessons of home electronics apply to VR: you don’t need a single vendor to make all your devices.
This mix-and-match ability is even more critical for enterprise customers. VR arcades, for instance, might use custom hardware or professional tracking systems. They want a software environment that is flexible and extensible. They want an environment that supports ‘off-the-shelf’ products yet extends for ‘custom’ designs.
OSVR already supports hundreds devices. Every month, device vendors, VR enthusiasts and the core OSVR team add new devices. Most OSVR plugins (extension modules) are open-sourced. Thus, it is often possible to use an existing plugin as baseline for a new one. With every new device, we come closer towards achieving universal device support.
A key OSVR goal is to create abstract device interfaces. This allows applications to work without regards to the particular device or technology choice. For example, head tracking can come from optical trackers or inertial ones. The option of a mix-and-match approach overcomes the risk of a single vendor lock-in; you don’t change your word processor when you buy a new printer, likewise, you shouldn’t have to change your applications when you get a new VR device.
We try to make it easy to add OSVR support to any device. We worked with several headset manufacturers to create plugins for their products. Others did this work themselves. Once such a plugin is ready, customers instantly gain access to all OSVR content. Many game engines—such as Unity, Unreal, and SteamVR—immediately support it.
The same is also true for input and output peripherals such as eye trackers and haptic devices. If developers use an API from one peripheral vendor, they need to learn a new API for each new device. If developers use the OSVR API, they don’t need to bother with vendor-specific interfaces.
I would love to see more enhancements to the abstract OSVR interfaces. They should reflect new capabilities, support new devices and integrate smart plugins.
More People Exposed to More VR Applications in More Places
Just a few years ago, the biggest VR-centric conference of the year had just 500 attendees. Most attendees had advanced computer science degrees. My company was one of about 10 presenting vendors. Today, you can experience a VR demo at a Best Buy. You can use a VR device on a roller coaster. With a $15 investment, you can turn your phone into a simple VR device.
In the past, to set up a VR system you had to be a geek with plenty of time. Now, ordinary people expect to do it with ease.
More than ever, businesses are experimenting with adopting VR. Applications that have always been the subject of dreams of are becoming practical. We see entertainment, therapy, home improvement, tourism, meditation, design and many other applications.
These businesses are discovering that different applications have different hardware and software requirements. A treadmill at home is not going to survive the intensive use at a gym. Likewise, a VR device designed for home use is not suitable for use in a high-traffic shopping mall. The computing and packaging requirements for these applications are different from use to use. Some accept a high-end gaming PC, while others prefer inexpensive Android machines. I expect to see the full gamut of hardware platforms and a wide variety of cost and packaging options.
“Any customer can have a car painted any color that he wants so long as it is black,” said Henry Ford. I’d like to see a different approach, one that encourages variety and customization.
On the hardware side, Sensics is designing many products that use OSVR components. For instance, our ‘Goggles for public VR’ use OSVR parts in a headset designed for use in amusement parks. We also help other companies use OSVR components inside their own packages. For those that want to design their own hardware, the OSVR headset is a good reference design.
On the software side, I would like to see OSVR expand to support more platforms. I’d like to see better Mac support and more complete coverage of Android and Linux platforms. I’d like to see VR work well on mid-range PCs and not limited to the newest graphics cards. This will lower the barriers to experience good VR and bring more people into the fold. I’d like to see device-specific optimizations to make the most of available capabilities. The OpenCV image processing library has optimizations for many processors. OSVR could follow a similar path.
Additionally, it is important to automate or at least simplify the end-user experience. Make it as close to plug-and-play as possible . The task of identifying available devices and configuring them should be quick and simple.
Simplicity is not limited to configuration. We’d like to see easier ways to choose, buy and deploy software.