It’s only been a few days since multiple publications started pointing us toward the possibility of an Apple car. (It’s so fresh that the phrase "Apple car" still feels a little odd to write.) In all likelihood, an actual car is years away from introduction — if Tim Cook and company decide to release it at all. Yet there’s already a rift growing in this young rumor mill: will Cupertino’s take on the future of transportation be capable of driving itself? The Wall Street Journal says no; Reuters says yes.
The truth is they’re both likely right.
The apparent disagreement in scoops from two of the world’s preeminent news sources speaks to a PR problem that has been plaguing the auto industry for years. "Self-driving" isn’t an atomic, yes-or-no feature; it’s a blanket phrase referring to an entire spectrum of capabilities. Some of those are on the road today and have been for years, others are still just a shade or two shy of science fiction. As Reuters notes, "Autonomous driving is likely to emerge progressively as driver assistance systems become more sophisticated."
"Self-driving" isn’t a yes-or-no feature
Even if an Apple car isn’t fully autonomous — something like Google’s vision, lacking even a steering wheel — it won’t ignore autonomous features wholesale.
In fact, by the time such a car is on the road, it may not legally be able to ignore them. Some features that fall under the purview of "self-driving" are so well established that the federal government has started recommending that automakers include them as standard equipment. So-called crash-imminent braking (CIB) and dynamic braking support (DBS) monitor the road ahead for situations that could cause a collision, and either automatically slow the vehicle down or help the driver apply additional braking to minimize the severity of a crash or prevent it altogether. For a brief, terrifying moment, the car is taking control of a core driving function on the driver’s behalf.
But beyond these pre-collision systems, convenience features on a wide variety of modern vehicles combine to closely resemble what you might consider to be "a self-driving car." A car with lane-keep and dynamic cruise control, for instance, can automatically stay in its highway lane and maintain a programmed distance behind the cars ahead of it, going all the way down to a standstill in the event of a traffic jam. Such systems are reliable enough nowadays that you can leave the driver’s seat altogether, as one YouTube poster did (unsafely and unwisely):
From there, it’s a small jump to a system that can safely change highway lanes without driver input. The Tesla Model S, with hardware enhancements included since late last year, can change lanes on its own by tapping the turn signal. Tesla has described its short-term goal for automated driving as "on-ramp to off-ramp," meaning that the car takes care of everything while you’re on a highway — it’s only on surface streets that you need to drive on your own.
Apple is notoriously deliberate about its approach to adding bells and whistles to its products; it’s rarely a first mover. (Just look at Apple Pay and Touch ID as recent examples — neither feature was first to market, but both are the best implementations in their class.) And no one is suggesting the Apple car is close to release — by all appearances, they just started working on it last year. That’s a blink of an eye on the glacial scale of automotive R&D. By the time an Apple car is on sale, it wouldn’t be viable in the marketplace without some manner of autonomous capability. It’d be akin to selling a car today without a USB port or Bluetooth connectivity, or selling a car in 2016 that doesn’t support either CarPlay or Android Auto. Yes, you might be able to do it, but unless you’re playing in the very bottom end of the market — a place Apple is not known to go — it won’t fly.
It would be a shock if Apple aimed for full autonomy with version one
Nor is WSJ wrong when it writes that "A self-driving car is not part of Apple’s current plan"; in all likelihood, it’s semantics. We’re still many years away from a production car that can reliably (and legally) navigate urban environments, crowded intersections, and complex city traffic patterns in all weather. It would be a shock if Apple aimed for that level of autonomy with version one, but it would be equally surprising if it came to market without the modern basics like pre-collision. Within a couple years, more advanced capabilities like dynamic cruise will become standard fare, too.
So yes, paradoxically, an Apple car can be self-driving and be not self-driving at the same time. Reuters can be right, WSJ can be right, everyone wins — except incumbent automakers, perhaps.