Why cars will never be smartphones on wheels
The digital transformation of the auto industry has hit a speed bump: Slower-than-expected electric vehicle adoption has delayed the rollout of intelligent, software-defined vehicles.
The big picture: Automakers are forecasting billions of dollars in recurring revenue from software and services that improve over time through constant updates.
That requires a next-generation electrical architecture — think of it as the car's brain — to handle everything from ride dynamics and safety functions to the in-car experience that consumers want.
Automakers dream of being more like Apple, with an elegant operating system like iOS that enables the same digital experience across all their vehicles.
Tesla and other EV startups already have such platforms, and they regularly send software updates to add features or improve performance.
But they had an advantage: Their modern cars were designed from scratch.
Legacy automakers are saddled with complex software networks cobbled together from more than 100 electronic control units that manage specific functions like braking or infotainment.
Even minor software updates are a hassle with such a fragmented system.
The shift to electric vehicles seemed like the ideal time for many automakers to toss out those antiquated architectures in favor of a Tesla-like approach.
Many companies who yoked software modernization to their EV development plans, however, are now thinking better of it.
Driving the news: Ford recently pulled the plug on an ambitious next-generation software project, deciding instead to pour those efforts into improving its existing architecture.
The fully networked vehicle project (FNV4) was to have been the foundational software platform for a future lineup of smart, connected vehicles.
In a blog post explaining the pivot, Doug Field, Ford's chief EV, digital and design officer, noted: "The world has changed since automakers, including Ford, laid out plans to rapidly redesign their vehicles for an electric future."
EV adoption has been slower than expected, he noted, and Ford's digital transformation shouldn't leave behind customers who prefer gasoline or hybrid vehicles, he wrote.
Instead, Ford will take a more incremental approach to software, building upon the digital experience launched recently in the Ford Explorer and Lincoln Nautilus and Navigator and deploying it across its full portfolio of vehicles.
The move will save money, Ford CEO Jim Farley told analysts this week, and even make future products more affordable.
Between the lines: The software conundrum demonstrates why a car is not a smartphone and Detroit will never be like Apple.
Phones are replaced every two or three years; cars stay on the road for 10 to 15 years or more.
That means automakers are often stuck spending money to keep old technologies alive, even as they're investing billions to develop more sophisticated, modern vehicles.
"Having a new electrical architecture on some vehicles and a legacy architecture on other vehicles and maintaining them all at the same time is not a good strategy for this new world," Field wrote.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Yahoo
28 minutes ago
- Yahoo
Robotaxi wars: How Waymo got the edge on the competition, and Tesla (so far)
'We are being super paranoid about safety." That's what Tesla CEO Elon Musk said after revealing the company's robotaxi test would start on June 22, later than expected, adding that the date could 'shift' beyond that. Meanwhile, betting site Polymarket lists only a 30% chance that Tesla (TSLA) launches the service before July, highlighting Musk's rosy predictions on self-driving that have been constantly pushed back. While Tesla is on the verge of starting a new chapter for the company, Alphabet's (GOOG) Waymo has been quietly racking up mile after mile, trip after trip, providing actual paid robotaxi rides to customers in San Francisco, Phoenix, Los Angeles, and Austin — where Tesla's test will begin. At stake is a huge opportunity for companies like Waymo, Tesla, and others. Goldman analyst Mark Delaney estimates the market for traditional ride-hailing in the US is worth approximately $58 billion currently, but could be worth more than $330 billion by 2030, with robotaxi-type services pushing the industry forward and reducing costs by — among other factors —not requiring a human driver. Waymo has an early lead, and it seems a big one. Rivals like Argo AI and GM-backed Cruise are gone, leaving only smaller players like Zoox (backed by Amazon), China's Pony AI (PONY) and WeRide (WRD), and of course Tesla. Waymo (before it was even called that) started off as an in-house startup as part of Google's X initiative way back in 2009. After some testing and securing of permits, the project officially became Waymo in 2016, and it started testing its Pacifica hybrid vans kitted out with cameras, LIDAR laser sensors, and radar. In 2018 Waymo and Jaguar paired up, using the British automaker's I-Pace EV for its testing; these are the vehicles most users are accustomed to seeing. Waymo's combination of vision, radar, and LIDAR, which the company calls its 'multimodal' approach, has the company claiming it has the best, and safest, robotaxi solution. Since 2020, anyone in a service area can download the Waymo app and hail a fully autonomous car. Waymo says it has the most robotaxi miles driven and that the company performs around 250,000 trips a week. A Waymo spokesperson said that the company has not yet seen a vision-only system that can achieve the levels of safety its current system has achieved, asserting that multiple sensors with overlapping fields of view are the best solution. Experts tend to agree. 'Waymo is undoubtedly the leader in autonomous driving technology, with their self-driving software (SDS) being widely regarded as best in class,' said University of San Francisco engineering professor William Riggs, an expert in autonomous technology. 'They have been around the longest, had patient capital and have the most miles on the road proving that self-driving tech works.' Riggs believes Waymo's software strategy combined with its diversified sensor suite, along with its manufacturer partnerships such as Chrysler, Hyundai, Zeekr, and Toyota, allowed Waymo to maintain a strong presence in the market and adapt to different vehicle platforms. Autonomous expert Katie Driggs-Campbell of University of Illinois' Grainger College of Engineering agrees. 'Initially, I believe their strength came from the computer vision/learning and compute expertise from Google,' Driggs-Campbell said, noting that Waymo has now 'blended real-world experience, recorded data, and simulation [generated] data to form a reliable system.' Waymo also touts its safety record, as well as how open it's been with data. 'We have an incredibly rigorous safety framework in place, and our safety record speaks for itself,' Waymo's spokesperson said, adding that the company has been voluntarily sharing safety performance data with the public, which goes beyond regulatory requirements. Waymo's approach could not be more different from Tesla's, and that has contributed to the dominant lead it sees currently. Tesla did not immediately respond to Yahoo Finance when seeking comment. But per Musk biographer Walter Isaacson, Musk was enamored with Tesla's vision-only, neural network-powered self-driving system when it was presented to him. He deemed it superior to Tesla's prior system, which ran on millions of lines of code using a rules-based system fed by data coming from radar and LIDAR sensors. Tesla's current FSD (full self-driving) and robotaxi software use only a vision-based system powered by a neural network 'brain' that is constantly training itself using videos collected by millions of Tesla vehicles. (And as opposed to Waymo, Tesla has reportedly asked the city of Austin to block release of safety records related to its robotaxi launch.) While Waymo holds a big early lead, what could stop the pioneer from growing more could be its operations and manufacturing ability. Riggs believes building vehicles at scale remains a significant hurdle for Waymo. 'Currently, they operate between 700 and 1,000 vehicles in their major markets and are producing roughly 1,500 vehicles annually. This relatively modest production volume could become a bottleneck for scaling their operations further,' he said. The other challenge for startups like Waymo in general versus Tesla is that their costs per vehicle are higher. Waymo's self-driving tech and sensors reportedly cost $100,000 plus the cost of the vehicle itself (for example, the Jaguar I-Pace retails for $73,875 in the US). While some of those costs have come down, scaling to new cities and providing enough vehicles to take on Uber, for example, will require more than the 1,500 vehicles currently produced a year. Tesla, meanwhile, has the ability to produce 2 million or more EVs a year, which include the hardware necessary for its FSD and robotaxi software. Tesla also has millions of vehicles on the road that could be converted for robotaxi use, though that would require the company's Hardware 4 autopilot software, which only started rolling out in early 2023. 'AI runs off of data. Waymo is trying to close that data gap with more sensors per vehicle; Tesla is looking to win with more vehicles with targeted data specific to their vision-only approach,' said Ken Johnston, vice president of data and analytics at tech consulting firm Envorso and former exec at Ford and Microsoft. 'Companies like BYD and Tesla also cannot be ruled out, as they continue to innovate and expand their capabilities in this space,' Riggs added. Xiaomi is also testing autonomous driving in China, but the company has suffered from safety issues in early trials. Johnson is bullish on Tesla's robotaxi offering, but the company faces a big challenge in surviving its safety test. Uber's self-driving unit and GM's Cruise could not overcome high-profile accidents. Testing the service without any accidents in a semi-urban environment is not easy. And unlike Waymo, Tesla is under federal investigation for its autonomous software. It also still needs to provide NHTSA with answers to how it plans to roll out its robotaxi program. Meanwhile, Waymo is plugging along, adding more cities to its list of current operations, with Atlanta, Miami, and Washington, D.C., coming online in 2026. Being backed by Alphabet helps too, giving Waymo capital runway for growth, investment, and long-term community outreach. Waymo's technological lead and strong backing from Alphabet clearly have the service in pole position. But this race is far from over. Pras Subramanian is the lead auto reporter for Yahoo Finance. You can follow him on X and on Instagram. Sign in to access your portfolio
CNN
an hour ago
- CNN
Musk fires up the robotaxi hype machine as his MAGA spotlight dims
Elon Musk has two key strengths: a bottomless well of money and an unmatched ability to hypnotize Wall Street with promises of a whizz-bang technological future. He's starting to run into the limits of both. On the hypnosis front, Musk is busy hyping the tentative launch on June 22 of Tesla's long-awaited robotaxi pilot program in Austin, Texas. Tentative, Musk said, because the company is being 'super paranoid about safety.' When even Musk says the date is iffy, do not hold your breath. The guy can deliver a game-changing EV, to be sure, but many of his more ambitious projects are notoriously behind schedule. In the meantime, Alphabet went ahead and began a taxi service using driverless Waymo cars in 2020, and now provides 250,000 paid rides a week in San Francisco, LA, Phoenix and Austin. (It is still a money-losing operation, sure, but the losses are essentially a rounding error for the tech behemoth.) If playing catch-up on robotaxis were Tesla's only problem, that'd still be a tall order. But the company is trying to usher in the future of transportation at the same time it is doing damage control on its reputation. What began as a kind of aspirational luxury car brand popular among well-to-do liberals has, for many, become synonymous with its erratic, authoritarian-adjacent CEO. Musk's MAGA turn wasn't the only reason Tesla sales started tanking globally, but it certainly didn't help. Chinese competitors have been eating Tesla's lunch in key markets around the world. The Cybertruck is — relative to Musk's original high-flying predictions — a flop, and the rest of the company's lineup is dated. Some disappointed drivers say they can't offload their Teslas because demand and resale prices have fallen so dramatically. In sum: Tesla, whose stock (TSLA) has tumbled 30% since its all-time high in December, needs its robotaxi event, whenever it happens, to really knock investors' socks off. Of course, in the past, Musk has been known to lean into the flash and somewhat gloss over the substance. Last fall, Tesla shares fell 9% the day after its 'Cybercab' unveiling, which offered few details about how Tesla plans to improve its still-unfinished 'Full Self-Driving' tech and featured robot staff that, unbeknownst to people who attended, were being operated remotely by humans. The Austin event will be a 'proof of concept exercise, not a rollout or a testing of a commercial asset for sale,' GLJ Research analyst and prominent Tesla critic Gordon Johnson said in a note this week. Part of the Musk mystique has to do with just how unfathomably rich he is — like, how could he possibly fail with all that wealth? And to be sure, Musk's money affords him a boundless budget to buy just about anything and sue just about anyone, in perpetuity. But Musk's foray into politics has shown that even Muskian wealth has its limits when Elon Musk is involved. For example, Musk dropped nearly $300 million on campaign spending for Donald Trump and other Republicans last year, only to blow up his relationship with the president a few months later with a series of outbursts on social media. He shelled out $44 billion on Twitter with the intent of building it into some kind of 'everything app' — an internet hub for commerce and a global town hall. The platform saw an exodus of users and advertisers as Musk took down speech guardrails, and it now more closely resembles the notorious message board 4chan. Musk's $20 million gamble on a conservative candidate for the Wisconsin Supreme Court was seen as a test of his newfound political power. But that one also backfired, putting Musk and his unpopular DOGE layoffs at the center of the narrative and enabling the liberal candidate to secure a 10-point victory. Bottom line: After torching Tesla's brand with his MAGA turn, Musk went and did something even more unthinkable and turned on Trump, straining (if not dooming) his status as 'first buddy.' Musk is now in damage control mode, and he's reviving an old playbook, revving up the self-driving hype machine, for a bit of redemption. 'He's got a problem,' President Trump told CNN's Dana Bash last week. 'The poor guy's got a problem.'
Forbes
an hour ago
- Forbes
While Tesla Talks, Waymo Drives
Tesla (NASDAQ:TSLA) has established itself as a frontrunner in the realm of autonomous driving for quite some time. Its Full Self-Driving (FSD) software improves the driving experience by automating specific tasks, enabling Tesla vehicles to navigate, change lanes, and park independently while necessitating the driver's active supervision. However, despite years of anticipation and high investor expectations, Tesla's FSD does not seemingly make a substantial contribution to the company's revenues, while other tech giants like Google have been ramping up their investments in self-driving technology. Concurrently, higher-margin software products such as FSD are becoming increasingly crucial to Tesla's growth narrative as its core EV business faces pressure from waning demand and heightened price competition. So, what is the current status of the FSD software? Tesla offers FSD as either a one-time purchase of $8,000 or a subscription priced at $99 per month. We believe the subscription option is likely to be the preferred choice for most new Tesla buyers due to its flexibility and lower initial cost. Although Tesla does not report FSD revenue separately, estimates from RBC Capital Markets indicate that roughly 5% of customers chose the subscription by early 2024. This attach rate has likely improved after Tesla reduced the monthly fee by 50% to $100 by April 2024. Based on vehicle delivery forecasts for 2025, we estimate that Tesla could sell approximately 1.7 million vehicles this year. If we assume that 10% of these customers opt for the FSD subscription, that results in about 170,000 additional subscriptions. This would raise cumulative subscriptions to around 460,000 by the end of this year. See detailed calculations in our dashboard How Big Is Tesla's Software Business? At a monthly rate of $100, that translates to approximately $550 million in annual revenue for Tesla. While this is significant, it remains a small portion of Tesla's total sales, and it falls short of justifying the substantial valuation premium that Tesla's stock commands on account of FSD's long-term potential. On a different note, if you're looking for upside with less volatility than individual stocks, the Trefis High Quality portfolio offers an alternative, having outperformed the S&P 500 and generated returns exceeding 91% since its inception. Elon Musk has continuously promised significant advancements in autonomy, yet Tesla has frequently missed its own timelines. For example, Musk had asserted that v13 of the FSD software would facilitate a 5 to 6 times increase in miles driven between disengagements compared to the v12.5 version. The miles between critical disengagements, which indicates the average distance a vehicle travels before the system must be disengaged, is regarded as a vital safety metric for FSD. Nonetheless, data now reveals that v13 yielded barely a 2.5 times improvement, increasing from 200 miles to about 495 miles. The head of FSD at Tesla has previously mentioned that for Tesla to achieve truly unsupervised self-driving, the average number of miles per critical intervention would need to match the 'equivalent of human miles between collisions,' which stands at approximately 700,000 miles. By this measure, it appears Tesla has a considerable distance to cover. Conversely, Alphabet's Waymo has been quietly and consistently making tangible strides in the real world. The company is currently conducting over 250,000 fully autonomous, paid rides each week across Phoenix, San Francisco, and Los Angeles, a significant increase from around 10,000 rides merely two years ago. Waymo's vehicles operate without any human oversight, in contrast to Tesla's FSD, which still requires driver supervision. Additionally, riders seem to have a preference for Waymo; according to Earnest Analytics, Waymo retains riders at a higher rate than Uber or Lyft. This implies that once individuals experience fully autonomous rides, they favor them. Safety could also be a compelling selling point. Waymo's safety report from last year indicated that its autonomous vehicles achieved a 78% reduction in injury-causing crashes compared to human drivers. See: Waymo To Separate From Google? That being said, Tesla has some advantages at this moment. Waymo's system relies on a comprehensive hardware stack that includes custom lidar and radar sensors as well as cameras that are retrofitted to the vehicles it procures from traditional automakers. In contrast, Tesla relies on a camera-only system powered by neural networks. This approach makes Tesla's system potentially simpler and cheaper to scale. Furthermore, there are likely over 5 million Teslas currently on the road, and it's reasonable to assume that a significant percentage of these vehicles are equipped with the necessary hardware to run FSD software. If the software can advance accordingly, Tesla could activate a vast FSD user base, swiftly altering the economics of its business.
