Rumors are continuing to trickle in about Apple’s long-expected car project. On Wednesday, CNBC reported that Apple is close to finalizing a deal for an Apple Car to be manufactured at Kia’s assembly plant in West Point, Georgia, an hour southwest of Atlanta. Apple is a famously secretive company, and I don’t have any inside information about Apple’s plans. But I’m skeptical that whatever product Apple ultimately unveils will match CNBC’s description of it.

According to CNBC, the Apple Car, due out in 2024 or 2025, will be “fully autonomous.” One source told CNBC that Apple is aiming to make “autonomous, electric vehicles designed to operate without a driver and focused on the last mile.” CNBC predicts that the cars could be used for food delivery or in a robotaxi service.

If true, that would represent a dramatic departure for Apple. A central feature of Apple’s corporate culture is that it sells hardware products directly to users. There are technology companies like IBM and Microsoft that focus on selling to business customers. There are technology companies—from Uber to Google—that focus on building services.

But while Apple does run some services and does of course sell products to businesses, neither activity has been Apple’s focus. Apple’s flagship products—the original Macintosh, the iPod, the iPhone, the iPad, the Apple Watch, and so forth—have always been designed as consumer-focused hardware products. It’s hard to believe that Apple, already taking a risky leap into a new industry, would simultaneously toss aside a core aspect of its business philosophy.

Robotaxi and delivery services would be a poor fit

Entering either the robotaxi or delivery markets would require Apple to do just that. Apple would either have to launch its own delivery or taxi service, or it would have to sell cars to partners that actually own cars and run services. Neither of these approaches would play to Apple’s strengths.

If Apple launched its own service, it would have to develop expertise in a wide range of new activities, from cleaning and repairing vehicles to dealing with local government officials. Given Apple’s scale, it would be very difficult to expand such a business quickly enough to have a meaningful impact on Apple’s bottom line.

Selling cars to partners would be an equally big departure. It’s true that Apple has made some products, like the Mac Pro, that are primarily purchased by business customers. But these products have always been peripheral to Apple’s business, and Apple sells the Mac Pro to individuals as well as corporate customers. By contrast, a delivery vehicle or robotaxi would probably need to be specifically designed for that purpose. Consumers likely wouldn’t even have the option to purchase one outright.

Apple has long prided itself on controlling the entire user experience. That has enabled Apple to maintain a high level of quality, earn strong customer loyalty, and charge a premium for its products. If Apple started making and selling Apple Cars to partners building taxi or delivery fleets, it would lose control over how the cars are used and would risk being blamed for the bad decisions of its partners.

So it seems more likely that Apple will build a conventional car that it sells directly to customers. And if it’s partnering with a major automaker like Kia, that suggests that the Apple Car will be a conventional, full-sized vehicle.

Apple has a lot to bring to the table here. Apple has deep expertise in battery technology that it could apply to the design of an electric vehicle. Apple could also apply its skills in user interface design to build the car industry’s best user experience.

Apple could follow in Tesla’s footsteps

The big open question is what Apple will do on the self-driving front. While CNBC says the Apple Car will be “fully autonomous,” it’s hard to believe that Apple is on a path to full autonomy by 2025—if “full autonomy” is even a meaningful concept.

In 2019, Apple only logged about 7,500 miles testing its vehicles on California roads. For comparison, Waymo, widely seen as the industry leader, tested its vehicles for more than 6 million miles in California, Arizona, and other states in 2019. And despite those efforts, Waymo’s self-driving taxi service is only available in a 50-square-mile corner of the Phoenix metropolitan area.

Of course, the Apple Car isn’t due out for another three to five years, and Apple might make significant progress during that time. But the company has a lot of catching up to do.

It’s hard to believe that Apple is on a path to full autonomy by 2025—if “full autonomy” is even a meaningful concept.

Apple’s best approach may be to follow in the footsteps of Tesla, which began shipping a home-grown driver assistance system in 2016. Initially, Autopilot’s capabilities were fairly limited, but Tesla CEO Elon Musk predicted that the system would improve rapidly. Musk claimed that the hardware was suitable for full autonomy (it wasn’t) and that the necessary software for full self-driving would be out in a couple of years (it wasn’t). In 2016, Tesla began charging thousands of dollars for a “full self-driving” software package that didn’t provide any functionality back then and still doesn’t provide the full autonomy Musk promised more than four years ago.

Apple obviously shouldn’t follow Tesla’s lead in making wildly unrealistic promises. But the basic strategy of selling hardware now and releasing software later might serve Apple well. The high cost of lidar in 2016 meant Tesla couldn’t afford to include it on every vehicle. The lack of lidar has made it more difficult for Tesla to improve Autopilot. By 2024, the earliest date an Apple Car might come to market, high-performance lidar is likely to be available for under $1,000. That’s cheap enough that Apple could make lidar sensors a standard feature.

Apple could use these sensors to initially offer a solid driver assistance system, then gradually upgrade the software over time to eventually enable fully driverless operation.

Apple’s design and battery expertise means the company may not need industry-leading self-driving software in order to ship a compelling product. The strength of Apple’s brand and design sense, combined with the inherent performance advantages of electric vehicles, should be enough to attract a lot of early car buyers. Customers may be happy to wait a few more years for more sophisticated self-driving features—especially if Tim Cook can avoid making unrealistic promises when the car is unveiled.

https://arstechnica.com/?p=1739884