Architecture and Virtual Reality: An Introduction and Evaluation of Hardware Options
Virtual reality, or VR, has experienced rapid growth within architecture, engineering and construction (AEC) industries over the past 3 years. What started as a curiosity has now become increasingly commonplace in architecture, real estate development, and construction. In fact, many building owners now request or require demonstration of virtual reality capabilities as part of the RFP process.
For building owners, VR provides a highly effective means of visualizing, marketing and selling an architectural project that consistently achieves significant return on the investment.
For construction companies, virtual reality dovetails with a BIM workflow and serves as an integral part of virtual design and construction (VDC).
In architectural practice, designers can rapidly prototype and more effectively convey architectural concepts throughout schematics and design development.
In all of the above cases, VR is an effective means of competitive differentiation, providing an edge to companies who are able to offer immersive virtual visualization and collaboration methods.
The goal of this article is to introduce the basics of what virtual reality is, and how it works, followed by an overview of the hardware types currently available. I’ll also describe some of the pros and cons for each hardware type, specifically as it relates to AEC industries.
What is Virtual Reality?
Virtual reality is a computer simulated environment, accessed through stereoscopic goggles that include a variety of different tracking mechanisms that track the viewer’s movement in physical space, and matches those movements within the simulated 3D environment. Headsets display a stereoscopic image, viewed through built-in lenses that creates the illusion of looking out into a space. 3D renderings of the environment are captured in ‘real-time’ – which means the computer is processing and presenting a new rendering of the scene 60 to 120 times per second. This provides the viewer with a very realistic sense of what it would be like to actually inhabit the simulated environment, which is what makes it such a powerful opportunity for architectural visualization.
Why Does it Matter for Building Design and Construction?
People have a very hard time visualizing architectural concepts. It’s nearly impossible for most people to fully understand what an architectural space will be like by viewing 2D floor plans, illustrations, animations or scale models. These are all ultimately abstract and distorted ways of visualizing a design, and not at all representative of how we experience architecture in the real world.
With virtual reality, we can gain a very holistic understanding of what the building will feel like before investing the massive amount of resources it takes to make it a reality. It’s a lot easier to change a pixel than it is to move a brick wall after it’s built. VR helps people understand architecture in a way nothing else comes close to, which we believe will ultimately lead to the creation of better, more efficient built environment. _
BIM -> VR
For architectural visualization applications, the simulated environment is most often created with a real-time engine like Unity3D, where a 3D model – often exported from Revit, SketchUp, or other BIM applications, is used as a basis for creating the virtual reality application. Materials and lighting are then introduced, and the virtual reality application is published as a file that anyone with the appropriate VR hardware can access and experience. _
The Importance of Framerate
Before we get too far ahead, I first want to make an important note about framerate, and why it’s vital to quality architectural visualization in virtual reality. In VR, framerate is defined as the number of times per second a ‘frame’ or rendered image is displayed on the screen, most often expressed in frames per second (FPS). If the framerate is low, or a number of other factors aren’t performing adequately, the viewer can experience ‘lag’ or ‘judder,’ where the simulated 3D environment they’re viewing isn’t properly mirroring their physical movement and rotation. This is one of the most significant and avoidable causes of motion sickness, or what is sometimes called ‘simulator sickness.’ _
One of the biggest contributors to low framerate, is when complex 3D models or very large graphics are being displayed, but the hardware system being used doesn’t have the capability of displaying the same at an adequate framerate. This can most often happen with hardware like Google Cardboard and Gear VR, as we’ll describe in greater detail below. Architectural models exported from software like Revit, ArchiCAD or SketchUp are notoriously complex when compared to ‘game ready’ models that have been prepared specifically around achieving high framerate. Professional VR developers employ numerous tricks to alleviate this issue, such as occlusion culling, baked lighting, remeshing complex geometry and more. _
If you are attempting to provide architectural visualization in virtual reality in any capacity, you absolutely must be sure to achieve a lag and judder-free experience. If you don’t, you will invariably cause motion sickness for many people who are trying VR for the first time, a perception that cannot be easily undone. _
Spectrum of VR Hardware Choices
Not all virtual reality hardware is created equal. They range from low cost / low quality, all the way to high cost / high quality. Each of these technologies has a place in architectural visualization. On the lower cost / high convenience / lower overall quality end of the spectrum, you’ll use your own cell phone as the display and depend on the phone’s internal rotational tracking sensors. On the higher cost / lower portability end you’ll need a dedicated, higher end PC to use it, and depend on the full capabilities of a dedicated GPU. This end of the spectrum also includes projection rooms, or CAVE systems. This post will not include an evaluation of projection-based systems, and will instead focus on the current generation of head-mounted display (HMD) options that will soon be available to consumer markets. _
HTC Vive
The HTC Vive is currently available to developers only, but pre-orders can be made starting in the end of February. The Vive uses lots of sensors including a MEMS gyroscope, accelerometer and laser position sensors, along with a Lighthouse ‘capture volume’ of approximately 15 feet by 15 feet (4.5 by 4.5 meters) . Photo sensors on handheld input devices enable it to be tracked by lasers mounted on opposite ends of the space.
Pros for architectural use: The ability to stand up and walk around in a space is huge for architectural visualization. The full 360 degree tracking effectively eliminates motion sickness, and provides an incredibly immersive experience. Even though the volume you can walk freely within is less than 5 meters by 5 meters, you’ll still feel a lot more immersed and naturally even standing still, since it so accurately tracks your position as you look and move around. Also, the input devices open new doors for architectural visualization and interactivity, enabling you to reach out and touch various interface elements. The ability to reach out to move and position furniture or other architectural elements will be a new frontier for VR.
Cons for architectural use: Not very portable. You can theoretically pack everything up and install it for demos on the road, but it will be cumbersome and time consuming. You’ll want to leave at least 1 hour for setup and take-down, and will have to decide on a case by case basis whether the hassle is worth it for your intended demo purpose. You’ll need to set up tripods for the lighthouse units that need to be installed above head-height, and the location where you’re demoing will need a large enough space to walk around in.
Oculus Rift
Almost all of our projects prior to consumer devices was built for the Oculus Rift Developer Kit 2 (DK2), which includes a motion tracking system with external camera that tracks infrared dots on the headset. This allows it to track your movement – duck down, or lean to the side and significantly alleviates motion sickness over the previous DK1 prototype or Gear VR that lack positional tracking. Unfortunately, the DK2 is no longer available for purchase directly from Oculus, but there are still some units available for purchase on sites like Ebay. 
Their new consumer version Oculus Rift (‘CV1’) is now shipping, and features an even more powerful positional tracking system called “Constellation” with infrared tracking sensors that track LED’s on the headset. The system is capable of room scale tracking, so the user can stand up and walk around within the space. The new Rift will have higher quality lenses with a wider field of view, and integrated headphones for 3D audio. Oculus is also releasing a pair of wireless, handheld controllers called Oculus Touch., that are fully tracked in 3D space – so you essentially see your hand in VR. To use the Rift, you’ll need a PC with decent GPU at least equivalent to a NVIDIA GeForce GTX 970 or AMD R9 290. You’ll also need a CPU at least equivalent to Intel i5-4590.
Pros for architectural use: Easy to setup, and much more portable than HTC Vive. Very comfortable, with built-in audio headsets and microphone. The ‘Touch’ input devices shipping later this year will also provide new opportunities for interaction that we’re looking forward to working with.
Cons for architectural use: Performance targets will be reaching 90 frames per second (fps) with the new Rift, which will require a very careful attention to optimization compared to the 75 fps target for DK2. This will be particularly troublesome for large and otherwise complex Revit conversions that haven’t been manually optimized, since it will likely result in a laggy experience that could cause motion sickness. For architectural applications, there may need to be renewed emphasis on ‘remeshing’ or optimizing environments, furnishings and fixtures to achieve this higher framerate. However, when it’s done right, the quality and sense of immersion is amazing.
Google Cardboard
Google Cardboard was developed by Google, and uses a folded cardboard holder with lenses. Cardboard not only uses the phone’s display, but unlike GearVR it also relies on the phone’s internal rotational tracking sensors. This makes the experience noticeably more choppy, and most users can only comfortably view applications for less than a minute or two. However, the cost and convenience of Cardboard are a huge advantage, making it ideal for certain types of applications.
Pros for architectural uses: Very convenient and portable – can be custom printed with your logo, images, etc. Very convenient and inexpensive. Serves as an effective gateway to higher end VR, with a lower barrier to entry for developers and users alike.
Cons for architectural uses: Limited primarily to static pre-rendered views, or relatively small environments with limited realism in lighting and material data. Much laggier than higher end systems, which leads to motion sickness much faster, and generally uncomfortable for extended use. Lack of positional tracking also contributes to motion sickness for some users, and because it depends on the phone’s on-board processing power, the amount of detail and realism that can be displayed in real-time walk-throughs is very limited.
Samsung Gear VR
Gear VR is developed in a collaboration between Samsung and Oculus VR. It’s different from Oculus Rift and HTC Vive in that it doesn’t require a high end PC to run it – but instead requires a Samsung phone that attaches to the front of the unit, making for an ‘untethered’ virtual reality device. It uses the phone’s processor, but augments the phone with an on-board rotational tracking sensor that’s more accurate than the sensors on the phone itself.
Pros for architectural use: Easily portable, for taking to clients without having to bring a high end PC. Better processing capabilities than Google Cardboard.
Cons for architectural use: Lack of positional tracking makes GearVR quite a bit less comfortable in terms of motion sickness for some users, particularly in architectural applications. Because it depends upon on-board sensors only, and lacks the processing capability of the more powerful PC graphics card, truly real-time virtual reality experiences displayed on a GearVR, other than static panoramas, and environments that have been heavily optimized (often cost prohibitive for architectural applications, will often be of a visual fidelity akin to what computer graphics looked like 8 or 10 years ago – plastic and flat, and often laggy since it struggles to achieve the framerate required to reduce motion sickness. GearVR also requires the use of specific Samsung phone types, and is prone to battery drainage and overheating when used for extended periods.
Portable PC
For the higher end virtual reality experiences achieved with the new Oculus Rift and HTC Vive, I described the lack of portability as a downside, which it certainly is. However, it’s worth mentioning that there are PC options available that have very small cases and are easy to fit in a carry-on suitcase. Our current favorite is the Tiki, by Falcon Northwest. This is the same machine Oculus has used for many of it’s own demos, and it packs a significant GPU punch with an NVIDIA 980.
Non-VR Applications
One often overlooked benefit of creating a real-time application for virtual reality with an engine like Unity is the ability to publish the same application as a non-VR version that those without VR hardware can access. Many times, it doesn’t require any extra effort at all – particularly if the application is designed with this flexibility in mind from the beginning.
Your Projects in Virtual Reality
Hiring Help. If you don’t have an in-house development team, the quickest way to bring your projects into virtual reality is to work with a VR developer to translate your architectural models into virtual reality. Similar to working with a traditional architectural illustrator, they can take your Revit, ArchiCAD, SketchUp, AutoCAD and other formats, and translate them into a virtual reality application. 
Arch Virtual has developed a multi-faceted workflow and toolset for importing and optimizing architectural models for use in virtual reality, and has already helped architects, construction companies, real estate developers and building owners all around the world visualize millions of square feet of pre-construction architectural design. We employ a combination of automation techniques to save time and development cost paired with manual optimization to ensure the highest possible quality and performance. We also have a team of developers trained specifically around this workflow, who can quickly translate architectural models into virtual reality experiences. To get an idea of the cost of our services, you can get a free estimate by telling us more about your project here:
Developing VR In-House with Immerse
Many architecture and construction companies have the team and technology to create visualizations in-house. This makes it quite a bit easier to utilize virtual reality as an integral part of their process. To assist individuals and in-house development teams with the process of creating VR applications, we’ve decided to make our own toolset available as a resource to save them considerable time and money and recently started a division of Arch Virtual called ‘Immerse Interactive,’ exclusively dedicated to developing these tools. 
Immerse Framework: Interactivity Plugin for Unity
Immerse Framework is our first toolset, currently available from the Immerse Interactive website (a division of Arch Virtual). It provides an easy to use toolset for bringing interactivity and a variety of comfortable navigation options that make it very easy for the end-user to explore large architectural spaces. 
Immerse Collaborative, Multi-Player VR
Immerse Collaborative is an easy way to introduce multi-player functionality to VR applications, which opens the door to deeper levels of communication and collaboration between project stakeholders. For Unity developers, it takes under 5 minutes to bring this functionality to a virtual reality application, and the login process for each user is extremely simple and hassle-free. 
Immerse Asset Library
Immerse Assets provide a library of triple-A quality 3D assets that can save the time and money it would take to develop these assets from scratch. They also provide visual consistency to projects that is hard to achieve by drawing from a diverse range of resources for each asset in a project. Immerse Assets will soon include residential, commercial and medical assets that can be purchased on the Unity Asset Store.
Conclusion
Virtual reality is on it’s way to becoming a commonplace technology in architecture, engineering and construction industries, simply because the value proposition is so clear. You can put on a pair of glasses that let you see into the future – to occupy a building before construction starts. The implications of this are staggering, and likely to disrupt the way we conceive of and visualize architecture. With millions of dollars, and a tremendous amount of resources on the line whenever we create new buildings, it’s important to get it right. The ability to extensively prototype and explore architectural designs in virtual reality will lead the way to more a efficient and effective built environment, ensuring that the building we create are perfectly suited to their intended purpose. 
