You’ve seen the video on YouTube: Steve Mahan, a visually impaired, conservatively dressed, white-haired gentleman with a walking cane and Ben Hogan-style hat on his head, climbs behind the wheel of one of Google’s nautical blue-colored self-driving Toyota Prius. He makes a “Run for the Border.” Then Mahan, who has lost 95% of his eyesight, picks up his dry cleaning, takes a joy ride around his middle class looking community while munching on tacos. Then he returns to his driveway, exits the car (tacos in hand), and heads off camera toward the house, safe and sound, never having touched a steering wheel, brake or accelerator.
The clip is amusing and tremendously uplifting—showing how autonomous driving technology can empower a person, giving them a sense of independence and freedom to do whatever they want, when they want to do it. Problem is, most Americans (disabled or not) can’t afford such a vehicle, even if one were available, let alone the advanced driving assist systems that are available to the public now, and serve as the building blocks of autonomous automobiles.
According to the National Automobile Dealers Association, the average American spends around $30,000 on a new car or light truck. In contrast, Interest.com’s 2013 Car Affordability Study says that the average American can only afford to spend $20,806 on a car. The featured Prius, which starts at around $24,000, is optioned up with a $75,000 to $80,000 Velodyne LIDAR system, visual and radar sensors estimated to cost about $10,000, and a nearly $200,000 GPS array. Not to mention the cost of the driving computer and software. Put into context: The staid-looking Toyota Prius Mahan “drove” around in the video costs more than a Ferrari 599. At $320,000, that’s an exclusive purchase, and well above the mean cost of a car, truck or SUV.
So, who, exactly are self-driving cars for? Proponents to make fantastic claims about autonomous vehicles and ADAS technologies, then automakers and other advanced safety system developers turn around and charge exorbitant prices for their creations, failing to factor in the need for mass adoption in the plan for success.
Fairness aside, cost will be an issue for driverless and self-driving car technologies well into the future. To provide all the vehicle electronic functionality, automakers must add complexity and unnecessary weight (wires, sensors and other components/modules) to a car. This can affect the cost of the vehicle, its performance, and how much it costs to keeping the vehicle on the road.
According to a recent study, “Emerging Technologies: Autonomous Cars—Not If, But When,” IHS Automotive forecasts that the price for the self-driving technology will add between $7,000 and $10,000 to a car’s sticker price in 2025, a figure that will drop to around $5,000 in 2030 and about $3,000 in 2035, the year when the report says most self-driving vehicles will be operated completely independent from a human occupant’s control.
“It’s a chicken and egg thing, the have’s versus the have not’s,” says John Absmeier, director for Delphi's Silicon Valley Innovation Center. “When people start buying the technology the cost will come down, but the cost has to come down before most people will buy it. Consequently, [SDC technology] will come in through the luxury, high-end market and penetrate down as economies of scale kick in.”
The incremental decreases in cost are projected based on the adoption of the technology (i.e., projected increases in sales of cars with SDC technology). IHS predicts that annual sales between 2025 and 2035 will jump from 230,000 to 11.8 million. That's about 9% of all the world's auto sales in 2035. Seven million of those 11.8 million vehicles will rely on a mix of driver input and autonomous control, with the remaining 4.8 million vehicles relying entirely on computers to get around. Combined with vehicles from previous model years, IHS also forecasts that there will be 54 million autonomous vehicles on the road by 2035. When will sales of autonomous cars outnumber those of conventional cars? IHS expects this tipping point to occur by 2050. By then, IHS says the majority of vehicles sold and those in use are likely to be autonomous, with conventional vehicles becoming increasingly rare.
Companies like Audi, BMW and Mercedes-Benz have a leg up on the competition in this case simply because their clientele is more affluent. “Our customers demand the latest technologies and are willing to pay for them, especially when it comes to safety technologies,” says Rupert Stadler, Chief Executive Officer and Chairman of Audi AG. “We are simply giving them what they want, and anticipating what the customer of tomorrow will demand from us.”
However, more mainstream brands, such as Ford, are looking more pragmatically at the bottom line. “We are not jumping into the self-driving arena simply to prove we can make a vehicle that can pilot itself under controlled situations,” says Paul Mascarenas, Chief Technical Officer and Vice President of Research and Advanced Engineering at Ford. “We know we can do that. We need to deliver it in the most efficient, cost-effective way possible, and still have an excellent user experience.”
The consensus in the automobile industry is that Google’s idealistic approach to the driverless car won't bring the price of these technologies down far enough in price to make its car a mass-market proposition. So most outfits are working on less exotic but much cheaper approaches to the driverless problem: They’re looking at ways to consolidate and simplify the hardware.
“We are looking into miniaturization, sensor fusion and integration of controllers, [to name a few technologies],” explains Ford’s Mascarenas. “Anything that will make the technology more accessible.”
Delphi’s has been working on a new type of controller that takes into consideration all three of Ford’s concerns. Its “multi-domain controller” is designed to take all of the data collected—from the body and security of the vehicle up through the active safety systems — and aggregate it into one controller, one processor that will then make the driving decisions and direct all the subsystems and sub-controllers in the vehicle. By combining all of the sensor inputs and processors into box, instead of hundreds, Delphi’s controller will bring down the weight of the vehicle (improving fuel efficiency and performance) as well as complexity of the system (making it easier to maintain.) In addition, it makes the vehicle cheaper to build; fewer ECUs and the corresponding wire harnesses to drive the costs up.
Another popular target for most ADAS developers is the sensor array. Tesla’s Elon Musk believes the LIDAR/RADAR sensor approach is too expensive. “It’s better to have an optical system, basically cameras with software that is able to figure out what’s going on just by looking at things,” he recently told Bloomberg.
Mobileye, a technology firm headquartered in the Netherlands, is aiming to get most of the functionality into cars for much less, with equipment that cost hundreds of dollars, rather than the tens of thousands. It currently focuses on helping cars avoid collisions and pedestrians, as well as drifting out of lanes – most of the advanced safety features currently rolling out in higher end autos. The longer-term goal, is more ambitious: a semiautonomous vehicle that can handle many, but not all, driving functions without driver input. To that end, it has developed a low-cost alternative to expensive sensor arrays that uses a single camera and a “system-on-chip.” It works as a third eye for the driver by supporting them in performing routine driving tasks (e.g., distance keeping, pedestrian identification, traffic-sign recognition) and provide timely warnings (e.g., lane departure warning, forward collision warning) in dangerous situations. Though he wouldn’t discuss exact prices has not been announced, Ziv Aviram, a co-founder and the company’s chief executive, told the New York Times that it is the “most cost-efficient system that’s out there.” One thing is for certain, the optical approach is currently more accessible, and Mr. Aviram is working with three powerhouses in the autonomous business: BMW, General Motors, and Volvo.
Driverless, self-driving, and autonomous driving technologies do have the potential to deliver improved safety and comfort for the driver and passengers, while reducing the vehicle’s impact on the environment around it and reducing the cost to the driver. But don’t expect to see a scene the like the one in Google’s video—where a blind neighbor drives by while munching a taco—anytime soon in your neighborhood. In reality, most people in Steve’s positionb who were head of household reported an annual household income of $25,550, according to the 2010 U.S. Census. Currently, the technology package in the Infiniti Q50 that includes adaptive cruise control, blind spot monitoring, lane keeping systems, and adaptive steering, costs an additional $6,600 above the base sticker price of around $37.000. Mercedes-Benz driver assist package costs around $3,000, but the base car costs just north of $92,000.
[Image: Flickr user Dave See]