Google has finally released a second Tango-compatible phone on the market: The ZenFone AR. The apps are fun to try out, but they also show limitations. And then there is that persistent competitor, Apple, which changes the game at the drop of a hat. What future does Tango have? How impressive is ARKit really? We investigated details about the ZenFone AR.
Tango is Google’s platform for augmented reality (AR) apps. Essentially, it’s about placing digital objects in the real world. These objects would then be visible on the smartphone display.
Tango uses 3D motion tracking and depth perception to allow the smartphone to know where it is within a room, so as you move around, these digital objects follow the movement of the phone and are thus seamlessly integrated into the environment.
Augmented reality can only be seen through a smartphone; some scenarios are still pipe dreams, but they may soon be real: Using an app, you can leave notes for your friends in a restaurant, or the servers could show you the dishes on the menu as a 3D animation. Naturally, digital shopping guides would also be a possibility with a smartphone app. Amazon may be planning to open AR furniture stores, for instance.
Now, Tango has made its second appearance in a consumer smartphone. The ZenFone AR shows the platform’s potential. We tried out some apps and games that are only available for Tango phones.
BMW i Visualiser: BMW comes to your living room
In a very brief period of time, BMW has developed a Tango app that places a BMW in your driveway, your living room or anywhere you please. The use case is not immediately clear at first, but it is quickly becoming apparent why augmented reality apps are so interesting. Because, when looking through the smartphone display, I suddenly see a BMW in front of me: either an i3 or i8. Stefan Biermann, Director of Sales Innovation, spoke to us about the BMW app’s background. His department has been working on the app for roughly a year.
But what’s the point? Biermann told us that he demonstrated a prototype of the app at a sales manager event. The added value of such a presentation became immediately apparent to all who were present. Finally, a dealer can use the i Visualiser to immediately show his customer different automobile color combinations. And he doesn’t even have to be at the dealership - the rendered car looks incredibly sleek, particularly with the ZenFone AR.
Wouldn’t virtual reality be a better alternative? Thanks to crafty tracking, a car can be integrated into any desired scenario. Biermann put a positive spin on it: Customers on sales floors would not feel good putting on VR glasses. On the other hand, an AR application is a whole other ball game: The smartphone here solely serves as a window into this imaginary world and the user—preferably the buyer in this case—always has control and has a feel for his environment. So, if BMW were to rely on VR platforms like the Vive, the manufacturer would probably have to provide an enclosed room for it. This makes the AR app more frugal; it only requires a square meter of free space.
However, an enormous amount of effort is required for the necessary 3D models. Even though BMW has the data, it must nevertheless be converted for the AR app and be implemented in the app. It currently requires a lot of manual labor, because the augmented reality development is still in the preliminary stages. Even the size of the three-dimensional models presents a problem. If BMW were to place the entire automotive range in the visualizer, it would quickly amount to several gigabytes.
The American Museum of Natural History resurrects dinosaurs
How can AR be used for educational purposes? The American Museum of Natural History has come up with an idea. Dinosaurs can be placed in the environment using the app. Unfortunately, the three-dimensional models are quite simple and not texturized, which reflects badly on the app. Nonetheless, the dinosaurs are animated a bit and move around somewhat. It’s a shame that the developers have not even managed to create a brief audio track with information on the ancient animals, let alone an interactive demo.
Do games always need to have a sophisticated story or even make sense at all? They certainly do not. Candy Trails is one such game. You start by first scanning the room, allowing Candy Trails to create the playing space. Done. Now the game is asking me to go up to a lollipop. OK. I look for the huge lollipop, go up to it and tap on the display. And then to the next lollipop. Well, the game seems to be quite boring. With the third lollipop, it finally makes sense to me: Because now I have to avoid walking into obstacles that suddenly appear as candy canes. The goal is to reach as many lollipops as possible. Candy Trails is definitely nothing more than a time waster, but a very interesting demonstration of the new gaming concepts that are possible with AR.
Creating large lines of domino pieces is not only time-consuming but also tedious: Every single domino piece requires your full concentration; otherwise, you them tip over and all your effort is worth diddly squat. Of course, that does not happen with an AR game. The domino game lets you create your own lines on the floor, use assorted color domino pieces, insert obstacles, or integrate special actions in the line. Unfortunately, there is no option for integrating real obstacles into the line.
The competition touts impressive stuff - using no sensors at all
As impressive as some Tango apps are: the competition is not slacking off. Two to three years ago, it seemed clear that AR Apps required hardware support à la Tango, but Facebook, and mainly Apple, have proven otherwise. For instance, the recently presented ARKit by Apple can precisely analyze the smartphone camera's image without using any additional hardware sensors at all. At the keynote, Apple presented an impressive spectacle on the table, which was visible with an ordinary iPad.
And the past several days has shown one thing: ARKit delivers on its promises. Many demos show impressive precision. And iOS 11 gives all iPhones 6S and higher, iPad Pro tablets and the latest iPad (fifth generation) full ARKit capabilities.
On the other hand, Google has most recently promoted the Tango platform, above all. It is very respectable from a technical standpoint, and yet, it faces a disadvantage: Smartphones have very thin margins, and each component costs extra. Will 50 percent of all Android smartphones really be equipped with Tango sensors in two years, as some people are imagining? I am inclined to doubt that due to razor-thin margins in Android smartphones.
In testing, Tango has shown that while it does do a decent job indeed, it doesn’t do a fantastic job. Yeah, Tango works precisely, but it actually can’t place the objects three-dimensionally in the environment. The augmented reality layer in all the apps that I tested remains an overlay over the camera image.
So, AR overlays reality where the AR content needs to be overlaid, but that doesn’t need to be excessively important. Because anyone who simply wants to hang little AR pictures on their wall doesn’t need to rack his or her brain about it. But what’s really the problem is the following: Other AR platforms can achieve this without additional AR sensors. For instance, take the app 1600, which creates a history lesson out of a dollar bill. All that without a single sensor.
Augmented reality in general is a very exciting field. Not everything is perfect yet, there mainly needs to be new usability concepts, because the more complex apps are not comfortable for the most part.
Google Tango in particular got off to a bad start. Facebook’s demonstrations at F8 came very close to Tango technologies and, by all accounts, Apple’s ARKit took it one step further. The problem: Apple has made augmented reality possible in software; on the other hand, Google only has a hardware solution. But hardware cannot be upgraded.
Augmented reality apps on Android are facing dark times. ARKit will already have an immense installed base, which is crucial for app developers. Tango is only available for a very exclusive circle of users. The prevalence alone will pull the rug out from under Tango’s feet in a very real way. Tango would have to provide significant added value in order to hold its own against a software solution.
The extent to which Google will continue developing its Tango platform remains to be seen. If Android managed to keep up with regard to augmented reality, then Google must develop an alternative to ARKit. I’m willing to bet that they are already diligently working on it at Google’s headquarters.
Software is Tango’s only hope
In any case, both ARKit and Tango selectively use the same sensor: ARKit knows the smartphone’s position in the room thanks to the position sensor and accelerometer, although Tango uses this, too. Google must also find a way to perform depth perception with the smartphone camera alone. It will be exciting to see how elaborate that actually is. A developer revealed to us that most Tango apps measure the room only once using the time-of-flight camera, but afterwards make do with the smartphone’s position sensors. In other words: Even Tango apps mostly work in a mode that resembles that of ARKit.
If Google manages to transform Tango into a software solution, then Google may have a chance to succeed. For the time being, however, the competition from Apple may leave it in the dust.
Google wants to expand Android by adding an AR platform, which is incidentally unwelcome news for the entire Android platform: Over the next several years, AR apps will be among the highlights of new releases. As of now, Android users will barely benefit from it, and Google is in a tight spot.
Have you seen some AR demos? What do you think of the topic? Let me know in the comments below!