We're back from our first trip to the Augmented Enterprise Summit, where we got to experience the XR world through the lens of the enterprise user. This is short summary of what we saw, and our high-level take on the current state of all of the different approaches that comprise XR: assisted, augmented, virtual, and mixed reality.
It seems that of large organizations who are typically slow to adopt new technologies, many are XR-curious, and a few are attempting highly-focused, limited-scope pilot programs led by internal innovation teams. A good number of smaller organizations and startups seem to be jumping in with both feet, hoping to gain first-mover advantage in a world where the technological geology is still forming and the floor is still lava, unclear what will solidify into stable ground and what will break off into the sea. It feels a bit like the late nineties, when big companies were all wondering "Do we really need a web presence?" That history has been written, and given the billions of dollars flowing in to XR R&D right now, it seems inevitable that something in this universe will stick. It's still the wild west, though, with lots of opportunity for talented and motivated developers with new ideas.
50% of the companies we talked to fell into one of the two camps mentioned above, with the remaining just now thinking about dipping a toe into these waters, concerned about missing the boat. That said, the event seemed heavily weighted towards big industrial segments - oil and gas, aerospace, defense contracting - and very sparse on the consumer brands. Most companies seemed interested in XR for assistance in manufacturing and true line-of-business applications, recognizing that there's not much of a consumer market to speak of, let alone target, just yet.
VR in the enterprise has taken hold, and the killer app here without a doubt is training and simulation. One attendee told me their company uses VR for training in "RIDE" situations - Rare, Inaccessible, Dangerous and Expensive. Judging from the number of exhibitors demonstrating training scenarios involving heavy equipment, heights, and things on fire, it was obvious that enterprise VR has found a niche in industrial scenarios. Most of these demos look like low-budget video games, though the simulations are undoubtedly more impactful than video-based training. All that said, they're still a long way from the uncanny valley. There's clearly a big opportunity for improvement, and we imagine this segment will grow quickly.
Meta's Quest 2 still can't be beat for the price, and given the simplistic graphical content in most if not all of the training simulations we saw, we can't imagine most companies would need more than this for their needs. Some companies may take umbrage at Meta's data security policies, so a thorough compliance review is recommended before moving forward here.
Pico's headsets are a serviceable and affordable alternative to Meta's offerings, but a lack of FTC approval makes them a non-starter for many organizations. Still, it's interesting to see China's strongest foray into the VR field with a solid contender.
VR's answer to mixed reality, Passthrough VR merges the physical world with the virtual via live camera feeds, often augmented with other sensors. This seems to be the fastest growing segment in the category. We saw passthrough in limited form in Meta's Quest 2. The Quest Pro and Quest 3, the VIVE XR Elite and the (yet-to-be-demoed-by-us) Apple Vision Pro all rely on this approach, which clearly removes some of the biggest hurdles faced by true mixed reality approaches.
The XR Elite latest has adjustable diopters, but wouldn't work with some of our team's prescriptions so things were blurry, and its passthrough feature had issues capturing and representing hands -- moving one's head from side to side, it was clear when it was switching between the cameras on either side of the headset, making hands jump a bit, ruining the illusion. Apparently Vive's primary selling point to the enterprise is, well, that they're not Meta -- they touted that they're the only XR headset to meet DoD's criteria for privacy and data security. The wireless streaming feature is solid though, meaning that if you have a truly demanding graphical application you can bypass the onboard GPU completely and offload the heavy lifting to a PC.
Meta's Quest 3 is an incredible step forward beyond the Quest 2. Its passthrough feature is the most compelling implementation we've seen to challenge holographic projection. Framerate was very high, the view was truly immersive, and while it was clear that you were viewing a rebroadcasted view of the real world, the simulation was nearly flawless, and its LIDAR-based hand tracking made for a truly solid experience. We were also impressed that the headset accommodated oversize eyeglass frames, although apparently through a partnership with Zenni you can get prescription optical inserts made for a sharp, glasses-free experience.
Sadly absent from the show was Varjo, who currently have the highest-resolution headset available. We imagine this would be useful in fields like architecture and engineering, once remote collaboration platforms become more robust. We look forward to trying out their latest XR-3 headset , although the high pricepoint and yearly subscription requirement would relegate it to highly specialized uses only. Also, while other headsets keep getting smaller and lighter, Varjo's headsets still appear bulky and heavy -- apparently that much resolution comes with both a high financial and physical pricetag.
AR's killer app seems to be remote assistance, giving remote team members the ability to assess, annotate and guide workers with boots on the ground via a shared first-person view. Here we see vendors leveraging off-the-shelf mobile devices (or those augmented with plug-on sensors), making it easy to deploy.
We did see our first glimpses of machine learning in these scenarios, enabling compute-assisted quality assistance in real-time -- in one demo, the interior of a real-world car bumper assembly was scanned, and computer vision was able to detect and point out items that were missing or out of alignment. Optical QA in manufacturing processes has been around for a while, but these AR-based systems are specifically designed to augment human vision, giving them superpowers to reduce error.
Assisted reality has been flying under the radar in the consumer XR world, but seems to have strong adoption in the enterprise. Our experience with RealWare's HMT-1headset was very unexpectedly the most compelling demo of the show. Essentially, it's a highly-ruggedized headset designed to give workers in industrial scenarios a hands-free second screen -- think of it as a tiny Android device floating just out of your primary field of view. Using voice commands (and only voice commands), wearers can quickly pull up and navigate images, schematics, or blueprints, share their live view and collaborate via Teams or Zoom, or simply turn on a (very expensive if that's all you're using it for) headlamp. Incredible sound cancellation lets you speak and be heard clearly on a videochat while in an incredibly loud environment (so we were told, this wasn't available to demo) Voice commands were easily mastered and highly responsive.
Its lack of long prophesied but yet-to-be-attained MR features like effortless gesture tracking or AR's real-time visual overlays are precisely what makes it so successful -- it does very few things, and does them incredibly well, making it useful and deployable today. Interestingly, it was also the most expensive headset on the expo floor.
Ah, the holy grail of the holographic heads-up display. The technology has come avery long way from the original Hololens we played with seven years ago. All that said, MR is still nascent and finicky, and like many emerging technologies we've seen, continues to be a hammer looking for a nail, particularly in the enterprise space.
The Magic Leap 2 was the only MR headset on the show floor. Magic Leap's latest was a standout for its ability to adjustably darken its lenses to counteract bright light, solving one of the key issues faced by the Hololens 2. It had a wider and more convincing field of vision than Hololens 2, and the externalized battery/processor worn on a sling makes for an incredibly light and comfortable headset to wear. Tracking was tight, and I was able to navigate a shared space with another operator (synced via the cloud) seamlessly. The demos we saw are apparently live products (from third party vendors), but neither were good demos of the full range of features possible with the Magic Leap hardware.
The Hololens 2 was absent from the show, but it's worth a mention here. Having recently deployed a public-facing experience that leveraged Hololens 2, we have good things to say about how far the product come from its first iteration. Gesture-tracking was much more solid, and the field of view was significantly improved. For the price, through, it feels more fragile and less visually impactful than the Magic Leap 2.
We have yet to see a compelling case for MR in the workforce. We still believe MR is great in shared, controlled environments that necessitate an awareness of the world around you. The opportunities to deliver a transformative experience via MR are great given its limited distribution (it feels like science fiction), but for now we're still waiting to see the killer app for MR that will dump rocket fuel on its adoption. In their keynote, Magic Leap talked about their excitement in the healthcare field, giving surgeons access to real-time vitals and a second pair of remote eyes on tricky surgeries, but we're not holding our breath for an MR set that can survive an autoclave.
There is a vibrant industry springing up around XR for the enterprise, but it still feels like very early days. Most manufacturers seem to be targeting consumers (where there's obviously a larger potential customer base) and gaming seems to be the killer app for most of these technologies. Beyond training/simulation and remote assistance, things like useful shared space collaboration and holopresence/telepresence still seem many years away. All that said, we continue to experiment with these technologies in order to best understand how to leverage them in meaningful ways on behalf of our clients.