Design for Spatial Computing
When designing computational systems at Oblong, we consider all the technology and people that inhabit a space. In a typical workspace or conference room, there may be several shared displays, laptops, mobile devices, cameras, input devices, servers, and, of course, people. Our spatial operating environment, g-speak™, makes it possible to construct meaningful relationships between the devices, displays, and people that share a physical space.
The One True Coordinate System
Most operating systems describe a display's pixel space using a 2-dimensional coordinate system with the origin (0,0) in the top left-hand corner of the screen. What does it mean to add a second display under these conditions? And what if that second display is at a ninety-degree angle to the first? Describing both displays in 2-dimensional space is no longer possible.
Oblong's g-speak uses a 3-dimensional coordinate system to describe displays and address pixels. We pick a central point, usually on the floor in the middle of the room, as the origin—our (0, 0, 0). From that origin, we can describe the location (and orientation) of any screen in space and create useful relationships between the different displays. This single, universal coordinate system also describes points beyond the boundaries of the screens, such as walls or surfaces, where no pixels are present at all.
Tracking Quality for Interaction Quality
The input devices that we design and build at Oblong are tracked and described—like the displays they interact with—in 3-dimensional space. Where other computing systems might know only the relative location of devices on a network, g-speak expresses the absolute position and orientation (the "pose") of an input device. This implies that you also know the attitude of each input device relative to a screen, or screens, because they live in the same conformed space. This is a qualitatively richer representation than the generic proximity descriptions of most consumer electronic devices. By virtue of this approach, pixel-less regions, as well as physical objects within the room, are also addressable by the interface and thus can become an integral part of the experience.
Space as Input
Because our input devices and displays occupy a single 3D coordinate system, it's possible to create natural interactions such as touch, hover, and free space gestures. Touch describes a physical connection with a screen. Hover describes an interaction near a screen, in which visual feedback depicts proximity to interface elements. And with an input device (a hand, a spatial wand) in free space, more expressive interactions become possible, including interactions that rely on or express spatial connection between user intent and the objects in a room. Chief among these, of course, is pointing.
Utilizing all three, the vocabulary for designing intuitive, spatial interactions is greatly expanded. No other system simultaneously enables pointing at a whiteboard and clicking to capture its current state; dragging content from one wall to another; proximally transferring information from a personal device to large shared display; or dropping a digital asset onto a physical device to print it.
As it happens, pointing is a central and crucial gesture. When a person points an input device or hand to interact with g-speak displays, other participants understand the area of focus through their inherent ability to know where the person is pointing. The gesture is thus simultaneously meaningful to the performer, to the observer, and to the digital system. Such shared geometric understanding is immensely powerful.
A spatial operating environment places valuable collaborative surfaces like whiteboards and tabletops on equal footing with their capable digital counterparts. In doing so, it pulls participants out of their personal screens back into a shared work space, encourages the flow of media and ideas across surfaces, and allows each participant to contribute and control the environment freely—all critical for collaborative work.
Into a Capable Future
A future built only of mobile computing isn't enough. Through their portability such devices are incredibly valuable, but this is offset by UIs that are reductive and incomplete. As devices and people become more mobile, we need an operating system that understands physical space—a fully capable, general purpose computing environment. This is what we've built: a system where all the critical characteristics of that UI (spatial, architectural-scale, collaborative, living) are built into g-speak as idioms and axioms. The resulting interactions are not only radically useful; they're expressive, powerful, and exhilarating.
Working with Watson
The goal of each Watson Experience Center—located in New York, San Francisco, and Cambridge—is to demystify AI and challenge visitor’s expectations through more tangible demonstrations of Watson technology. Visitors are guided through a series of narratives and data interfaces, each grounded in IBM’s current capabilities in machine learning and AI. These sit alongside a host of Mezzanine rooms where participants further collaborate to build solutions together.
The process for creating each experience begins with dynamic, collaborative research. Subject matter experts take members of the design and engineering teams through real-world scenarios—disaster response, financial crimes investigation, oil and gas management, product research, world news analysis—where we identify and test applicable data sets. From there, we move our ideas quickly to scale.
Accessibility to the immersive pixel canvas for everyone involved is key to the process. Designers must be able to see their ideas outside of the confines of 15″ laptops and prescriptive software. Utilizing tools tuned for rapid iteration at scale, our capable team of designers, data artists, and engineers work side-by-side to envision and define each experience. The result is more than a polished marketing narrative; it's an active interface that allows the exploration of data with accurate demonstrations of Watson’s capabilities—one that customers can see themselves in.
Under the Hood
Underlying the digital canvas is a robust spatial operating environment, g‑speak, which allows our team to position real data in a true spatial context. Every data point within the system, and even the UI itself, is defined in real world coordinates (measured in millimeters, not pixels). Gestures, directional pointing, and proximity to screens help us create interfaces that more closely understand user intent and more effectively humanize the UI.
This award-nominated collaboration with IBM is prototyped and developed at scale at Oblong’s headquarters in Los Angeles as well as IBM’s Immersive AI Lab in Austin. While these spaces are typically invite-only, IBM is increasingly open to sharing the content and the unique design ideas that drive its success with the public. This November, during Austin Design Week, IBM will host a tour of their Watson Immersive AI Lab, including live demonstrations of the work and a Q&A session with leaders from the creative team.
Can't make it to Austin? Contact our Solutions team for a glimpse of our vision of the future at our headquarters in the Arts District in Los Angeles.