At Mobile World Congress in Barcelona, Technical Fellow Alex Kipman (best title ever for a fellow who has a tech-related job) mounted the stage to announce that after four years, Microsoft had a successor to the HoloLens. HoloLens 2 is here, and it’s less of a consumer device than ever. That’s not a bad thing.
Next-gen improvements include a flip-up visor, lighter weight, more centered balance, and a new fit system that purports to make wearing the headset for long periods of time more comfortable. On the more technical front, the new version has double the field of view of the original, a new time-of-flight sensor, and new eye-tracking sensors. Microsoft says you can also now interact with holograms more naturally and instinctually, just like you would with physical objects in the real world. That capability is borne out of the fourth generation of Microsoft’s Kinect technology in tandem with AI tools, taking the fun gesture-recognition gameplay technology and making it truly useful.
More interesting in terms of real-world usefulness, Microsoft has developed ways you can easily make guides for training using Dynamics 365. For example, if you want to show a new employee how to repair an engine, you can overlay directions on top of the physical device to show things as nuanced as which specific tool you need and how to use it. Creating the guide, though, is the cool part; just pick various instruction items from a palette and drag and drop it into the guide. Then you can hand a new worker a HoloLens 2, pull up the guide they need, and set them on the task.
Although the HoloLens 2 works offline, it’s really designed as an intelligent edge device that leverages the Azure cloud. Using the new Azure Spatial Anchors, you can create “holograms that persist in a specific physical space.” The example Microsoft gave is that a clothing store manager could attach holographic images of outfits attached to mannequins that need to be dressed. “The next day, an employee could walk in, point an iPhone at each mannequin, see how it should be dressed and begin pulling clothes,” reads a Microsoft blog post.
You can also make 3D digital models with the cloud-based Azure Remote Rendering service. The idea is that instead of crafting physical prototypes or architectural miniatures, you can create them digitally and share them holographically through HoloLens 2. Because they’re digital, you can send the models to anyone who has a headset anywhere in the world.
The new hardware costs a cool $3,500, which is $500 more than the original HoloLens cost at launch. You’ll also need to cough up some subscription fees to use the aforementioned cloud tools.
You’ll note that none of the above appeals to consumers. It’s not supposed to. More than anything, HoloLens 2 shows, with great clarity, that consumers are not in the Windows Mixed Reality equation. The original charming Skype and living room-fun mixed reality demos are long gone. How did we get here?
A tool, not a toy
Microsoft has been known to bungle big projects, like the horrific Windows RT not-OS and its inexplicable murder of Windows Phone (err, Windows Mobile). When it announced this wacky HoloLens thing at an event in January 2015, which I didn’t attend for reasons that are lost to time, I simply did not believe the early reports of its greatness. There was no way that Microsoft managed to create a device that solved (or at least ably addressed) all of the issues that had plagued the new and exciting world of XR.
I’m actually kind of embarrassed by the article I wrote about it. The heart of my conclusion is this line: “Are we to believe that one Microsoft team has solved all the design, engineering, UI, input, processing, and content problems that an entire emerging industry is still struggling with?” Microsoft had, to an extent. At the very least, the HoloLens was a huge leap forward on all fronts.
I still remember vividly when I first had the chance to demo Microsoft’s HoloLens. It was a little later that year, at Microsoft Build 2015. It was a whole secretive operation, staffed by dozens if not hundreds of young people in blue Microsoft shirts. We were invited to a location and directed up and down elevator floors, having our IPD measured here, forced to put our cameras and laptops into lockers there, whisked into demo rooms and just as quickly ushered back out, and before we knew it we happy few were all back on the sidewalk and headed back to the convention center.
It was an astonishing set of demos. I mean, it was all there: holographic imagery that interacted with the real world, easy-to-use hand and voice controls, an untethered headset, tracking, etc. I sat down in the press room immediately afterwards and wrote it all down. I don’t know how long I worked, but after I hit the publish button and stood up to leave, the only people left in Moscone West were me and the cleaning crew.
What I don’t think I fully realized at the time—along with many others—was that the HoloLens was not a consumer device. I’m not sure Microsoft did either, at least not fully. At the time, it was right there in the mix with exciting VR devices in the Oculus Rift and HTC Vive, but it was always something completely different. Whereas the game demos we kept seeing on Rifts and Vive elicited squeals of delight and gasps of awe (at least from me), the games I saw on HoloLens were boring at best. There were endless demos for industrial and educational applications, which in hindsight I realize I should have valued more, but one key demo made me understand what the HoloLens is and what’s is capable of.
That happened at Microsoft Build 2016, a year after my first go with HoloLens. The “Destination: Mars” demo took place in a temporary big black box in the lower-level showfloor at Moscone West. Inside, a holographic Buzz Aldrin taught us about the red planet and threw it to a hologram of an actual scientist, who told us more, and then we got to watch a facsimile of the Curiosity Mars rover do something, and then we were invited to look at the barren surface and peak under virtual rocks. It was cool stuff and a really engaging experience, but it wasn’t until afterwards, that I realized what I’d just seen. I just had some follow up questions but ended up asking them of someone from JPL—an actual rocket scientist. He waved over another rocket scientist, and we talked for what seemed like an hour.
These are the most important things I learned: Those rocks? They were not some artistic recreation of Mars rocks; we were looking at close-up, 3D photos of actual rocks that are actually on the planet’s surface. If there was a space bug crawling on the rock, the resolution was so good that we could have counted its thoraxes. “Do you think someday you could use this technology for mission-critical applications?” I asked. “We already do,” replied the rocket scientists. Every day, these guys show up to work, and while sipping their morning coffee, they review the images that Curiosity sent home overnight. The images are mapped onto accurate 3D frames and added to the treasure trove of intergalactic data that the rover captured the day before. And there I was, in a very terrestrial building, wearing a headset smaller than a motorcycle helmet, looking at the Mars the same way rocket scientists get to see it.
I realized, finally, and clearly: HoloLens is a tool, not a toy. The launch of HoloLens 2 takes that to a logical end, rolling out complementary technologies and applications that leverage Microsoft’s strengths to create a package that the company believes is ready to deploy into the workplace.
I still believe that AR has an enormous role to play in the consumer space. Give me a pair of spectacles that are about the same size and weight as the ones I’m currently wearing, and make them a super-peripheral for my phone that gives me a HUD, endless monitor space, BT keyboard and mouse support, a way to interact with my environment, and reliable comms, and you can have all the money you want. We just aren’t there yet. In the meantime, we have what looks like a mature suite of industrial, medical, and educational XR tools from Microsoft, four years after the HoloLens was originally unveiled.
Obviously there are limitations, cost being chief among them. At $3,500 a pop, companies can’t just casually nab fleets of these things. It’s a huge investment, and the only way they’ll bite is if HoloLens 2 can reduce costs in the aggregate. ROI is king. Microsoft certainly appears to believe that’s what the new package can offer.