Antony Vitillo

@SkarredGhost

Why I’m not completely sold on the ARKit revolution

We’re living the ARKit hype era: everyone is talking about it, a lot of people are experimenting with it and there is this opinion that ARKit is the next big thing and that it will disrupt completely the AR and VR worlds. I think that ARKit is a great technology, but honestly I don’t share all this hype and I’ll tell you why.

Last Saturday I was chatting with a friend of mine on a bench. She started asking me something about the VR field and I told her that in autumn there could be a disruption thanks to Apple ARKit.

She: “What is it?”

Me: “Think about Pokemon Go, the moment when you shoot at Pokemons… like that, but that works. You can see virtual elements on the real world”

She: “Ah, why?”

Me: “Well, for instance to play. You can see little creatures in the streets and shoot at them”

She: “Ah, cool. But…why?”

The dialogues went on with various “Why”s to which I couldn’t provide a satisfying answer.

The standard user wants a reason to use a tech. As innovators we like using new techs because they’re cool, but usually people want something more. There is a reason if the sentence “Content is the king” has become so popular. If ARKit will be able to provide valuable content at its launch, it will become ultra-successful. Otherwise it may also be a failure.

The problem is… I’ve still not seen a convincing ARKit demo. I’m not talking about the mythological “Killer app”, I’m talking about something worth using.

Most of ARKit demos on the web are “I’ve put a 3D model on a plane in the real world”. Cool, eh, ARKit is able to add lighting and shadows and that’s incredible… but after the first 2–3 that I’ve seen, I’ve started getting bored by that.

Then there are the “This is how we’ll do things in the future thanks to ARKit” demos. My favourite ones, the ones that just make a prototype and without any practical reasonings think that they’ve invented the future.

Let’s take for instance this

Very nice demo! But… how do you know the location of all items in all supermarkets? How do you keep these locations updated when the employees move the elements? How do you convince all the supermarket brands to give their maps to you and not making each a proprietary app for that purpose? Furthermore… is it interesting for supermarket owners to take you directly to your products? (The more you stay inside the shop, the more you buy!)
I had a similar idea years ago with my buddy Gianni. We then discarded it because of all the practical difficulties.

Then there is this great ARKit maps project

It is surely a cool demo. But… do we really need to see those arrows in front of our eyes? Let’s consider that in order to see them we have to walk with our phone in our hands, in front of our eyes, looking like zombies. Imagine all people walking like that in the streets as during the Pokemon Go fever. Is it practical? No. It is more practical to see the map on Google Maps (or Apple Maps), so you can just take a glimpse to your phone, zoom in, zoom out and knowing what to do in the next minutes, not only in the next block.

Then there are the marketing apps. The brand X has made an ARKit app to amaze their users! (sarcastic)Wow! In this sense, ARKit will be the new 360 video. I expect lots of marketing booths and experiences showcasing ARKit stuff just because ARKit gives the “Wow” effect, as they’ve made with 360 videos. The problem is that they’ll exploit the technology so much that in the end no one will be excited by it anymore, exactly as it is happened for 360 videos.

Then there are the last demos: VR positional tracking made with ARKit

Apple will make possible 6DOF mobile headsets! Yes… maybe. I mean, I haven’t tried this solution myself and can’t tell you, but I have some little doubts:

  • How does the battery last if the phone does VR+ARKit?
  • Has ARKit the necessary precision and framerate for nausea-free VR? In AR, if a virtual element trembles a bit, this is not a problem. But if in VR you perform the tracking with a non absolute precision and speed, you cause motion sickness. When using HoloLens, I’ve seen sometimes the holograms moving a bit and HL has far more cameras than Apple phones.

Can’t speak about the iPhone8 (that will probably have depth cameras and other cool stuff), but surely current phones CAN’T perform such kind of positional tracking with a single RGB camera. Look for instance a tracking comparison with Tango phones

I know, you’ve seen experiments, but they’re just experiments. In my past startup, we performed positional tracking on mobile devices using Kinects. It was cool, but people after using it reported dizziness and discomfort due to Kinect low precision and framerate.

I repeat: iPhone 8 should solve at least part of this problems thanks to multiple depth cameras, but I would like to see if tracking will be good enough for positional tracking… without draining too much battery.

Then there is the business side, that convinces me even the least. I mean, when talking about VR and AR now I hear people telling me: leave VR, because AR with ARKit is the future! When I ask them why, the answer is always the same: “500 million devices by the end of the year! A huge market!”. No one is interested by the tech, everyone is interested only by the market and hence the possible money. But, let me be honest: this is not an answer: selling potatoes would give me even a bigger market, but if my interest is in making immersive applications, then farming is not for me.

Jumping on a technology just because of the big market has little sense. At first one has to identify the problem to be solved and then solve it with the appropriate tech and not vice-versa. If I want to develop a training experience, for instance, most probably VR is the tech to go. If I want to make movie streaming, a 2D app is the most suitable one.

All this hype in my opinion is inflating itself. And I’m afraid this bubble can pop when ARKit will be out, exactly as happened with smartwatches. Surely ARKit will make people go wow, because every iPhone owner will try all the available demos for the first days… but… after that? If after that there won’t be valid applications, he/she will stop using AR and then we’ll hear everyone telling that AR is a fad. I’m tired of it.

Let’s stop inflating this bubble and let’s start giving AR a reasonable value. In my opinion ARKit is cool because it is an enabling technology, that let people enjoy phone-based AR in an easy way, that let people do new things. But it has some big problems:

  • It requires you to see through a phone. This means having to put constantly the phone in front of your eyes and look like an idiot. Furthermore, AR through a screen is less magical than AR through glasses like HoloLens;
  • It doesn’t let you operate hands free. In the shoppping example, if with one hand you hold the phone and with the other the shopping basket… with which hand are you going to take groceries? The map example with the arrows would be incredibly useful when you’re driving… but you can’t drive while holding a phone inside your hand!
  • It is friction-ful. Whenever you want to experience AR, you have to take your phone out and point it towards whatever. Friction is always an issue. Glass-based AR is better because it can be contextual: you are walking on the street and suddenly you see an augmentation on a building that invites you to enter because your profile is compatible with your activity. With ARKit YOU should decide when to see AR elements and this is bad.

This problems will be solved with time and with the advent of AR glasses… and we all know that ARKit is part of an Apple plan to arrive to glasses. But we’re still not there, there still a long road to go.

Honestly, I don’t think that ARKit will disrupt the way people will use phones. I think that people will continue use the phones as know, but will also have the ability to use augmented reality for some well tailored applications, where the use of an AR augmentation through the phone has sense. For instance, interior design is a very practical application of it and in fact IKEA is planning to develop an ARKit app. Other companies are making similar applications, too.

Another example is the above AR map app: the only moment when I would use it is when I’m about to reach my destination. Google Maps is great in taking me there, but then sometimes I don’t get the exact place of the destination… having an augmented arrow telling me “it is here” would help. Then for instance there is this misuration app

What I want to mean is that IMHO ARKit is a great technology and has some interesting applications and will surely offer new opportunities to iPhone owners. Furthermore I see it as incredibly useful to spread the word about AR and VR to a lot of users in the world. But I still don’t see it as a complete disruption of the market, something that will change our lives forever. I work with phone AR since years and I’ve never been excited by it that much. Maybe I’ll be proven wrong (I hope so, since I love AR), but I admit I’m still not competely sold on it.

Of course I would like to know your personal opinion about it and open a debate about this topic. What do you think?

(Whatever the answer, if you like what I write please join me at my AR/VR blog The Ghost Howls!)

More by Antony Vitillo

Topics of interest

More Related Stories