USAToday Column

August 10, 2015
The future of cars is smart, not autonomous

By Bob O'Donnell

FOSTER CITY, Calif.—Americans love their cars (and trucks). They also love technology gadgets. So, it’s not the least bit surprising to see strong interest in efforts to merge the smart technology of our gadgets into our cars.

For some, the penultimate expression of this combination is the autonomous, self-driving car. Efforts by Google, as well as car makers like Audi and Tesla, have received enormous amounts of press and fostered speculation of highways full of machine-driven cars by the end of the decade.

The reality, however, is likely to be far different. It’s not that the technology isn’t there to make cars that can drive on their own — it clearly is. But the practical, legislative, and insurance requirements that are going to surround the widespread usage of autonomous cars are likely to keep them from becoming mainstream for a decade or even longer. The problem isn’t necessarily with a single car, it’s about getting a lot of cars from a variety of vendors all working together in a coordinated fashion. That is not a trivial task.

What we will see, moving forward, is increasingly smart cars. Those smarts are going to come at many levels — from the relatively simple but still important notion of better integrating our smartphones and popular mobile applications into our cars’ in-vehicle infotainment (IVI) systems, to better car connectivity, to increasingly sophisticated advanced driver assistance systems (ADAS).

Today’s cars already offer a surprisingly wide but not always well-known array of ADAS systems, from lane drifting warnings, to collision detection systems, to 360-degree camera views to performance enhancing adjustments to drivetrains and suspensions, and more. Even good ol’ cruise control — arguably one of the first ADAS features — is evolving to the point where it can do some basic autonomous driving of its own.

Not surprisingly, there are a lot of technology companies involved in making these kinds of improvements and their automotive business is growing. Companies like graphics giant nVIDIA, who reported a 76% increase in their automotive business during last week’s earnings report, not only can power the graphical display on car infotainment systems, but they’ve also created chips specifically designed to read, react and learn from sensors and cameras built into cars.

Similarly, modem maker Qualcomm is often providing multiple radios to car makers — one for the car’s telematics systems (such as OnStar) that can report safety or accident data automatically, another for entertainment and mapping system updates, and yet another for in-car WiFi.

Another technology company, ARM, known for its semiconductor design and licensing, literally and figuratively drives a number of today’s car subsystems: climate controls, back-up cameras, powertrain adjustments and much more.

As great as these new technologies in cars are, however, they do come with a potential downside, as the highly publicized recent Jeep “car hack” clearly illustrated.

It turns out that traditional car design commonly connects many of the various components to a main bus that runs throughout a car. As a result, finding a “break-in” point in one system can lead to potential problems in another. It’s obviously a big security issue for which car makers are using yet more technologies — such as an embedded security chip offered by companies like SIM card giant Gemalto — to help address.

Another challenge is how cars are currently designed and manufactured. Unlike the 6-month design cycles of tech products, most car products are on 5-year plans, so today’s newest component technologies won’t be arriving until model year 2019 cars.

From a technological perspective, that’s why I think it’s going to be much longer before we start seeing a mass of autonomous cars. But the problems for them go well beyond technology. The fact is, many people like to drive.

Sure, commuting in rush-hour traffic can get old, but cars have long held a much more important role than a simple here-to-there device. Cars represent freedom for many, and it isn’t until the thought of that freedom going away starts to hit, that you really appreciate the concept.

Plus, let’s be honest. There are enormous issues surrounding machine control versus human control that we will soon be facing in our society, and autonomous cars represent a very personal instance of those upcoming battles. Safety is certainly an issue, and there’s no denying that some level of automated driving could potentially help save thousands of lives.

However, it strikes me that the path to that level of safety is best served (and more realistically achieved) through continued refinements to our existing cars — making them smarter on many levels.

While I look forward to enhancements to my driving experience, I’m not yet ready to let go of my steering wheel.

Here's a link to the original column: http://www.usatoday.com/story/tech/columnist/2015/08/10/future-cars-smart-not-autonomous/31411613/