A Dedicated Driver Bets on Connected Car

It’s 6 o’clock in Santa Clara and I inch onto a jammed Highway 101.

I like to call this my favorite drive, because I finish in my own garage, where my kids await. But tonight, as on most evenings, the commute is mostly slog.

The fancy wheels crawling alongside me are a far cry from the first car I called my own: way back when, I bought my aunt’s battered blue ʼ89 Camry with the dead radio, then drove it into the ground.

I may never own one of these Ferraris or Porsche 911s I see piloted by Silicon Valley’s best and brightest, but I appreciate great cars and great driving. And I’ve gotten a lot more interested in cars the past several years as cutting-edge software assumed a greater role in the driving experience.

As a technology guy dedicated to securing the new digital world, I want to enhance the drive. I want to inject safety, convenience and security without subtracting control, freedom or passion.

A wave of digital vehicle enhancements is about to break over us. Connected vehicles will become the norm. There are naturally risks to work on along with the benefits. But there is more for driving enthusiasts to cheer here than you’d think from reading the news.

Cars Already Computers on Wheels

Even inexpensive cars have long come loaded with small computers – electronic control units, or ECUs. There are 50 to 100 in a typical vehicle, assigned to control airbags, steering, brakes, diagnostics, you name it.

There will be vast safety and operability benefits when they’re connected to a central onboard intelligence center and vehicle-to-vehicle communications systems. It’s difficult work, because of component diversity and complexity, and because most existing components weren’t designed to “converse” digitally – in fact, some still-used designs predate the consumer Internet.

But when that’s solved, cars will “read” roads and give drivers better data. They will trade messages about their routes and rates of travel, helping to head off accidents. With an autopilot, I can take calls while I’m stuck on the 101.

Of course there are cybersecurity risks. Yes, a skilled hacking team remote-hijacked the controls of a Jeep Cherokee a few months ago and sent it hurtling into a ditch, and the story drew a lot of headlines.

The attackers used the Jeep’s in-vehicle infotainment software as an attack vector, then exploited a buffer overflow; connected with an old-school device called a controller area network bus, or CAN bus; and modified one of the car’s many ECUs. They sent their disruptive messages via that CAN bus.

But I’m confident about improved vehicle cybersecurity. What you didn’t read about the Jeep hack: Technology exists that mostly could have frustrated it. The current challenges are elderly technology within vehicles – the Cherokee’s CAN bus, like those in many cars, was a vintage 1980s part – and the sheer number of attack vectors a typical car presents.

All human interfaces on a car, and every technological “way in” from wireless key fobs to onboard-diagnostics ports, need security checks for malformed data, then additional firewalls.

Secure Ecosystem a Must

There are about 35 major vehicle brands in the U.S. selling nearly 300 models, so this gets complicated in a hurry. We need to see common technology standards and what the experts call a system-level cybersecurity architecture approach – simply, more seamless armor presenting fewer attack opportunities.

Toward this end, Intel Security is backing an Automotive Technology Review Board to investigate and establish best practices. The federal government also is working with the car companies on standards-based seamless V2V communication. We want to see a common, secure technology ecosystem for future cars, supported by automakers, device manufacturers and security vendors.

The car industry is onboard. It’s spending billions in the next few years developing autonomous-car technology. Tesla stock is rising on speculation the company will commit to introducing fleets of shared driverless cars as soon as 2018. IHS Automotive predicts annual sales of autonomous cars will approach 12 million by 2035.

Everything really hinges, however, on getting that secure ecosystem right. If the technology isn’t stable and secure, the public shouldn’t be asked to trust it.

Here I’m inspired by the evolution of other technologies born before the Internet was a thing. Take Microsoft’s Windows operating system. The first Windows release was in the pre-Web 1980s; subsequent generations suffered Internet malware attacks around the turn of the century. But coders responded and today Windows 10 is a creature of the connected world, much closer to impregnable.

Automotive cybersecurity is on a similar trajectory. We’re doing a global upgrade of automotive technology born when the only connectivity concern was pulling in a decent FM station. We’ll get there. But for us to succeed and earn trust, automotive suppliers must take a hard look at repurposing components unchanged since before the Berlin Wall fell and the Pet Shop Boys were in heavy FM rotation.

The next generation of cars has to earn affection as well.

I vote no to a sci-fi future where we have to stow our open-road yearnings and be shunted along slotted roads inside people-mover bubbles that have no steering wheels. If there’s a misconception about autonomous-car technology, it’s the idea that it will irrevocably curtail a driver’s authority.

The movement really is about empowering the driver – to assign certain tasks to onboard intelligence while remaining in command. Choice is key. I want my autopilot when I’m grinding through a Bay Area rush hour, but I’ll stow it to storm Route 100 in Vermont, the Blue Ridge Parkway in Virginia or the Pacific Coast Highway at sunrise.

Most of us have favorite drives we still want to tackle with our bare hands. But if the computer and its sensors see and act on things the human brain misses because of fatigue, repetition or blindspots, we get safer personal transportation in the bargain.

The secure, trustworthy, connected car is about to change driving. Hazards will diminish. Annoyances will recede. We will be safer.

But the poetry of a short-throw gearbox, beckoning 2-lane blacktop and a tach begging to be redlined always will beckon drivers who want it.

I don’t see how we change that. I don’t see why we’d want to. This long, slow Highway 101 commute, though – that’s another story.

Christopher D. Young is senior vice present and general manager of Intel Security, an organization committed to protecting digital experiences. Follow him on Twitter: @youngdchris.