Apple's Project Titan Explored
For a hot minute, I considered titling this post Apple's Project Titan(ic), but beyond that being far too click-baity of a headline, its also entirely unfair to any project, especially one of which we know so little. You can't go a week in the tech blogosphere without hearing of some project reset, a high profile leader departing / being poached from a competitor, or a presumptive company death-knell if Apple doesn't ship something next month. All of these rumors keep floating around, showing that there is clearly some smoke here, but is there fire or a smoke machine?
This got me thinking... what exactly is Titan? Its become short-hand for "Apple's autonomous vehicle", and while that's likely something resembling the truth, what I haven't seen anywhere is anyone think thru how Apple of today (and really, Apple of 5-10 years ago when this strategy was likely conceived) is putting in place the foundation pieces to be able to actually build a vehicle. Since I hadn't seen it, I thought, why couldn't I put that together, so whenever this mythical vehicle comes to light, I can look back and see how foggy my crystal ball was.
An aside.. What's in a name?
Project Titan. The Titans are the 12 pre-Olympian gods, who are themselves children of primordials. They are the parents of the Olympian gods who we have been part of the human mythology for around 3,000 years. After being overthrown, the titans dewelled in a realm of murky gloom from which they could not escape. I'm not suggesting that the same fate is in store for Apple's project, but I could see, given its little known history, how some might draw that parallel.
Or could it be named after Saturn's largest moon, Titan. This moon is an interesting parallel to Earth, in that it has a significant atmosphere and stable bodies of liquid on its surface. While these liquids are far different than our water-rich Earth, the moon Titan is considered a likely alternative source of life in our solar system.
There was previously an automobile company named Saturn, originally started by former General Motors executives, and then purchased by GM and incorporated as its own division in the larger company. It ceased production in 2009, and was disbanded in 2010. I consider it highly unlikely that Apple is in any way modeling the Titan project around the Saturn, and it definitely isn't taking design queues from the company. It is interesting to note, however, that Saturn did have some unique, at the time, business models, such as no haggle pricing, which is something that fits with Apple's go to market approach very well.
Apple the company tends to invest in core technologies that can be used across its many platforms (macOS, iOS, tvOS, watchOS, etc), which are then exposed to customers in different ways. Consider that Apple's smart assistant, Siri, is available on just about every Apple platform, but the experience of using it on the TV is very different from using it on a HomePod. As I think about the technologies that would make up Project Titan, I feel that much of what Apple is doing is building core technologies for its current platforms, with an eye to how it could leverage those same technologies on some type of autonomous vehicle platform. Lets take a look at what those core technologies are and how they might be leveraged by an automobile.
coreOS. The core operating system for all Apple's platforms is very similar. Apple builds in such a way that the code can be written once but work on many types of hardware. The core operating system for a low-power device like the watch shares a significant amount of code with their highest end iPads & Macs. You have to think that Apple is building any new platform to leverage this huge investment that they have already made.
Apple Silicon. It started as CPUs for their phones & iPads, then it expanded to GPUs, to processors for its laptops and desktops. Apple acquired Intel's cellular modem division and is currently thought to be working on designing and building their own modems. The processors Apple makes are seen as some of the fastest and most efficient in the industry, far outpacing competitors such as Intel, AMD & Qualcomm. Their lead only keeps getting bigger year over year.
How Apple achieves this lead is multifaceted, but for the sake of this discussion, lets focus on a couple areas. First, because they build the hardware and software, they can build exactly and only what they need to make their offering better than the competitions. Tesla has recognized this and has moved to making their own silicon for their vehicles as well. Second, to do autonomous well, you need very fast parallel processing, and a lot of it. Apple has shown with the architecture of their M1, M1 Pro and M1 Max processors that they can do this better than anyone else at this time. But there are more reasons...
Machine Learning. Dig a little further into Apple's chip architecture and you'll see that they aren't just making CPU & GPUs, they're adding their own machine learning chips into the mix. Early on, these types of chips were too specialized and costly, both to build and to run, so that they existed mostly in data centers. As the technology has improved, it has made its way into more and more devices. These components can be designed to do everything from facial recognition in your photos, to recommendations on what notifications to allow during the work day, to real-time augmented reality. It isn't a stretch to think that an autonomous vehicle will need this to be able to do things like make turns, avoid obstacles on the road, etc, and ML chips will be deeply involved in making those decisions.
Augmented Reality. Apple hasn't released much in the way of products that take advantage of this technology yet, but its been coming for years now, and is currently on their 5th generation of software to make AR apps a reality. I can see a way where the vehicle's machine learning identifies road debris and then that debris is highlighted on the windshield for the driver to see it is in the path of the car.
Vision. Apple has some of the best cameras and camera software in the business. Their low-light camera software is incredibly good. Their ability to stitch images from multiple cameras together in real time is second to none. Take a vehicle, put these cameras all around it, pair that with their machine learning and custom silicon, and suddenly, you have a imaging supercomputer on wheels. Imagine being able to take a 360 degree video of your surroundings when on a drive up Highway 1 in California, and then playing that back later on your AppleTV. Add to that, in early 2020, Apple shipped its first product with LiDAR, and you can see that they're filling out the product suite needed for autonomous vehicles.
Sensors. Your iPhone, iPad and Apple Watch all have a tremendous number of sensors in them, measuring things like your location, heart rate, motion, altitude and so many more. Expect that any vehicle will be loaded down with sensors in every nook and cranny. Seats, dashboard, trunk, batteries, motors... the list will seem endless. The vehicle will monitor itself in ways unexpected but greatly appreciated by its passengers, even if they never directly see the work of those sensors.
Maps. Admittedly, their launch of their own mapping tools in 2012 were not without flaws, but it has grown by leeps and bounds in the years since. No, it still isn't the best in the world, but it closes the gaps with each release. It costs literally billions of dollars a year to build and maintain maps at the level Apple does, so finding new ways to leverage this investment makes it easier to justify the spend. They started partnering with others, particularly DuckDuckGo search engine, but being able to use maps in a vehicle they produce is an even bigger opportunity.
Privacy & Security. This may seem an odd one to include, but I think its even more imparitive with a vehicle than a cell phone. Think about the amount of data a vehicle will collect, not just on where the driver goes, but on everything that happens around it. If someone gains access to your phone, that's bad, but if a bad actor were able to take over your vehicle while you are at speed would potentially be much worse. Apple's investments in building safe & secure platforms will be invaluable to future platforms they build.
Music & Apps. No one can deny the impact that the iTunes music store and the App stores have had to how and how much we use our devices. Expect that these will be extended to any vehicle Apple makes. It will have great sound. It will be enhanced by apps built by 3rd parties, to make your driving experience even better. You'll have access to your entire music library; no more need for terrestrial radio stations or DJs.
Siri. I touched on this earlier, but when you think about Siri's potential for the car, especially for performing tasks that are historically dangerous while driving, could make the road much safer. If drivers are not digging for their phones, but able to hear and respond to text messages just like they do when conversing with a passenger in the seat next to them, that makes the road better for everyone.
Wallet. Apple recently announced adding keys and drivers license to the Wallet. Both of these things have been historically needed to drive a vehicle; things you had to carry with you. If they're not just part of your phone, but part of your car, then they are always with you. Your car could pay for your meal as you drove thru the drive-thru at your local fast-food restaurant.
Family sharing. How do you choose who gets to drive a vehicle? Typically you're ok with anyone of legal age in your family driving the vehicle. If you're all set up as a family in Apple's systems, you've already got your key and your license set up to drive the vehicle. Get in and go. Parents can see where the kids have gone, how fast they went and thru the use of "Screen Time", could only allow the vehicle to drive home after curfew.
Manufacturing partnerships & supply chain. Apple doesn't just design great devices, they make partnerships to build those devices. They own the supply chain inputs to ensure top quality and keep stock on their shelves. Apple doesn't employ the people who do the work of assembling their devices, but they do own the machines that those people use. Expect that Apple will build not just the vehicles, but also the manufacturing capability to produce those vehicles.
iPhone integration. I called this iPhone, but really, its integration with all of Apple's other platforms. I've touched on this in other sections, but its worth a section of its own. Knowing your phone is your car key, that the music you're listening to in the house transfers seamlessly when you get in the car, that the car knows where your dinner reservation is... these things lessen the cognitive load for we humans. When you're living in an all-Apple ecosystem, things just work, far more than their competitors.
That's the technology stack that could be leveraged into an autonomous vehicle. Its a pretty big list, but likely there is something that I'm missing. The list focuses on elements within a vehicle, but there are likely many supporting server-side technologies that Apple will also leverage to make the platform even better.
Some Lessons from the Past
Part of the smoke we see when journalists talk about Project Titan comes from the lack of credible information we have from inside Apple about what exactly it is that they are doing. Apple is typically closed lipped about their plans, but we can learn some things about what they've done in the past.
In the early 2000s, it was a joke that Apple had become the iPod Company; that their Mac division was a fraction of their revenue from music and devices. Everyone was expecting Apple to do the easy thing and to extend the iPod platform to just include a phone app. It seemed like this was what was going to happen when, in 2005, Motorola released the ROKR which had iTunes built into it. Everyone looked at that phone and thought, here's the first version of the iPhone. A few months later and everyone was shocked at the launch of the actual iPhone in that it was nothing like the ROKR. The iPhone was already functionally designed at the point the ROKR was announced, so Apple likely didn't learn a great deal from that partnership with Moto, but they didn't do it for no reason, either. Apple isn't the first company to jump into a new space; they typically jump in when they know how to do something better than anyone else has in the space.
That leads me to the point that Apple has already been in the car space for a while, with its CarPlay solution. Yes, this is a far from a full vehicle platform, but I expect that Apple has learned a lot about what is important to drivers and passengers thru this platform. These learnings likely will make their way into any eventual vehicle.
When the iPhone was launched, it wasn't what people expected. No one thought it would ditch physical keys for a fully-glass screen. Apple tends to look at what fits the need the best, not what people are expecting. Sometimes, like with the iPhone, this works incredibly well. Sometimes, like the initial Home Pod, it doesn't work nearly as well. I keep referring to Project Titan producing a vehicle and not a car, because the more I think about it, the less I feel like a car, in the traditional sense, is what is happening here. I think about that long list of technologies earlier in this article, and I think its more like Apple is building yet another platform which will be moldable into a different set of form factors. Just as Apple has phones that go from the SE thru to the Pro Max, I feel as if what they're building here isn't a single car, but more a platform for building vehicles, where the design and job to be done will not be what we are expecting.
If Apple up-ends our expectations on what a vehicle can look like, that also means that we're going to have trouble trying to figure out how we value any product that comes to market. The initial iPhone was thought of as ridiculously expensive, as it was not carrier subsidized, when all other phones were. People thought they were buying a phone, but didn't realize that what they were buying was a mobile computing device, which happened to include the ability to place phone calls. The iPad was just the opposite. Rumors prior to launch were that the device would be well inexcess of $1,000, despite it having much less functionality than a typical PC. The iPad launched at a price of $500, and was considered to not necessarily be a bargain, but to have a significantly higher value than anticipated.
During the launch presentation of the iPhone, Apple announced that it wanted to capture 1% of the cell phone market. Looking back, that is a laughable underestimate for what the iPhone would go on to become. If you consider all smartphones since then have largely followed in Apple's lead, then 100% of the market is now owned by the design paradigms that Apple instituted.
The first iPhone, iPad, Apple Watch and AppleTV devices all had massive deficiencies at launch. The iPhone had slow cellular network speeds. The iPad was hampered due to a paltry 128Mb of ram. The first Watch had anemic battery performance. The first AppleTV had incredibly poor software. Yet, each in their own ways, every one of these platforms has grown into a modest to blockbuster success. Apple usually gets the right concept the first time, even if that particular implementation is limited in some significant way. The vehicle will likely be the same. I won't begin to speculate what that limitation will be (not autonomous at launch, poor range, etc), but there will be something.
When Apple launches into a new category, their initial sales are relatively modest. Lots of first adopters are very excited, and the device may even be difficult to obtain, but that often is due to limits in the supply more so than an overwhelming demand, as Apple works to build a market for the product. Not every product is an iPhone, with exponential growth. Most products are linear growth, and for something as complicated to manufacturer as a vehicle will be, with a very long supply chain, scaling from producting 1 vehicle to 1,000 to 1,000,000 will not be an insignificant undertaking. Apple has the best teams in the world at this, but the challenges they experience will all be new as the category will be new to them. Don't expect them to outsell Tesla, or much less Ford, GM or VW Group, without many incredibly hard years of scaling hell.
I'm not a psychic, nor do I pretend to be one. I see patterns, I think about how those relate to one another, and make guestimates on where things could go. Its an art; it isn't science. Take nothing of what follows as what Apple will or should do, but as what one path to success might be.
First and foremost, Apple is a product company. They make products they sell; they don't build platforms for other people to repackage. They will work with a partner to build the product, but it will be unapologetically an Apple product. They don't put their logo on something that doesn't meet their standards. If you hear a rumor of Apple building a platform for a different auto manufacturer to put their logo on, you can forget that ever happening. If Project Titan releases a product, it will be an Apple product and no one else's.
Apple designs and controls their core technologies. They are often amazing stewards and provide these tools for others to use, such as Webkit, Clang, etc, but at the end of the day, they own it. Part of their history is that when they depend upon others, they are disappointed. Look for much of what is built to be influenced by the good and bad that has come before, but every element of the vehicle will be touched by Apple in some way.
Any product coming out of Project Titan will deeply integrate technology from across Apple's disciplines. They will leverage and extend the technologies they have built in other product lines to make any vehicle feel familiar, something that their customers know and love, before they ever sit down in it.
Our vehicles perform many jobs: they get us from point A to B, they haul our stuff, and they do all this with a high level of safety, both to us and others. Expect that any vehicle will be adaptive, so that it can change and fit itself to our needs, in ways that no prior vehicle ever has. This is the same philosophy as to why the iPhone went with a software keyboard instead of a hardware one. Once a hardware keyboard is produced, it is forever limited by its physical nature, where a keyboard defined in software can be updated infinitely. This approach will be a key to Apple's future for Project Titan.
Apple operates stores where it 1) sells its product 2) teaches you how to use the product and 3) fixes your product that was broken (likely) by you. What their stores do not do is maintain their products. You don't take your iPhone in for a tune-up every 3,000 hours of usage. Apple has no interest in making money off of servicing the product they sold you, so expect that any product they sell will need minimal to no maintenance, other than connecting it to wifi and power. Anything else is outside of the job Apple wants to do.
Legacy automotive manufacturers make vehicles obsolete well in advance of production beginning, as the next generation of vehicles is already in design years ahead of time. Tesla has started to show the older manufacturers what is possible if you update your models over time, providing new value to existing customers. Expect to see Apple not just follow this, but to take it in new directions as well. The first vehicle will likely ship with equipment disabled by default, or with limited functionality enabled, and that as the software catches up, will provide new capabilities even years after it rolled off the production line.
There is likely more; there must be more. I'm just a person who ponders these things in their spare time. I'm not paid to produce a vehicle platform, I just ponder what one might be. Apple has people smarter than me, many of them, in fact, who spend years thinking thru these exact same questions. If I got even 50% of the way to what they release, I'm surprised, but expect that I'll follow up here to let you know what I got right and what I did not.