Smarter Cars, But How Smart?

With the emergence of the Internet of Things, smart cars are beginning to garner more attention — the kind that comes with real R&D dollars, market development plans and cost analyses for future commercialization.

Smart cars are different than connected cars, which are simply smartphones on wheels. Until now, the focus on intelligence in automobiles has largely been on driver assist and information delivery—the GPS, OnStar, LoJack, among others—as well as infotainment, and integration with the smartphone. But all of that is about to change as the IoT evolves.

Courtesy of ABI Research.

Intelligent transportation systems (ITS) have been around for years. UC Berkeley has been at the forefront of this with what is called the Partners for Advanced Transportation Technology (PATH) program since the mid-1980s. In the early 1990s, Minnesota had a pilot program with driverless snowplows. There have been other as well, but they are all test or beta deployments.

Even though the technology has been on the evolutionary fast track, integration has been slow. There are a number of reasons for that. One is that there has been resistance from the automobile industry. While carmakers support semi-autonomous vehicles (driver assist), a fully autonomous transportation system will significantly diminish the revenue stream from vehicle sales.

“For car manufacturers, it is a little tricky to accept driverless cars because it disrupts their fundamental business model,” says Dominique Bonte, vice president and practice director at ABI Research. “They will no longer sell to private individuals, but to specialized organizations that distribute and maintain these vehicles.”

Such an infrastructure will simply be an on-demand environment. No one will “own” a vehicle. “Private resources will evolve to shared resources, centrally controlled, since autonomous vehicles can be controlled remotely,” Bonte says. “One can do traffic management, remotely control speed for optimal flow, for example, eliminating gridlock. ” There will be a plethora of electric vehicles everywhere. If one is needed, the nearest one will simply respond to the request, much like elevators do in high-occupancy buildings.

Another huge challenge is the motorway infrastructure. Major work will be required. Sensors will have to be attached to everything, everywhere, not just the cars – signs, signals, bridges, tunnels, crossings, among others. And, if not integrated into the roadway itself, they will have to be placed alongside the road or overhead. In any event, that will be a huge undertaking and the number of sensors will be voluminous.

A third issue, and without a doubt, the most significant is safety. With vehicles, there is zero margin for error. The smart car and the intelligent infrastructure have to be on and working every second. There cannot be even the slightest chance that corona mass ejection (CME), a terrorist act, power outage, or even a simply glitch in the network, can cause a fault in the system.

Finally, some less tangible issues and ones that are more nebulous to assess, revolve around public perception and understanding of what smart cars are, or will be. There are a couple of vectors that relate to this, comfort zone and mindset. Simply put, the general population today isn’t comfortable with hands and feet off the controls. Also, for some, there is a deep love affair with cars and driving – some of us are as attached to them as we are to members of our family – and that hasn’t changed much since the first Model T rolled off the assembly line. That kind of conditioning is difficult to alter. Plus, there is something gratifying to wheeling 450 horses to 60 MPH in less than five seconds (at least, to this editor).

“Generally, the industry experts believe that eventually both the mindset and the safety issues will be addressed and we will see some sort of autonomous vehicular system,” says Steven Shladover, professor and the former deputy director and current program manager for mobility at PATH. “However, don’t expect to see it in full ubiquity and autonomy until the next century.”

That may be a long time to wait considering that “basically, the technology for a self-driving car exists, more or less, today,” says Drue Freeman, senior vice president of global sales and marketing for NXP, pointing to the Google driverless car. “It relies on very, very expensive sensing technology to be able to do that reliably, but it demonstrates that such technology does exist.”

So, all of that being said, let’s take a look at what the state of this industry is today, and how it will play in the .

The apps
Smart cars will offer a number of benefits to the rider, among them, smart navigation. And that’s not just about route planning. It involves vehicle-to-vehicle (V2V) real-time communications, including route assessment, route dynamics (road conditions, traffic jams, road work, etc.), and dynamic rerouting based on this real-time data. There also will be much more sophisticated voice recognition, including algorithm-based voice print verification for security.

Courtesy of NXP.

The vision for smart cars is that they will provide most, if not all, of the routines we do while driving — individualized entertainment, automatic parking spot location, connecting to your social media, even car optimized, real-time carpooling, via the IoT. They will take on the role of Car as a Service (CaaS) rather than unique, individualized objects.

Tomorrow’s vehicle sophistication will rival that of the human body, which will be saturated with wearable sensors. In the vehicle, the driver and vehicle will become one. It will be able to monitor your vital functions, or other conditions such as heart and diabetes. It will be able to take action if an emergency arises such as a heart attack or stroke, pull over while alerting first responders and directing them to the scene. It will connect to your personal network and monitor schedules, routines, and tasks and automatically execute tasks such as picking up kids from school or making appointments.

Perhaps the biggest benefits would come from being green, and advanced safety features. Autonomous vehicles would optimize fuel efficiency and environmental emissions via optimal driving techniques. Functions like braking, turning and accelerating can be done with respect to other vehicles, minimizing the conditions that lead to accidents (following too close, cornering too fast, etc.).

But for all this to happen, there will need to be a “perfect storm” of collusion. “In order to fully unlock the automotive IoT potential, it will be critical to address a wide range of barriers including security, safety, regulation, lack of cross industry standards, widely varying industry dynamics and lifecycles and limited initial addressable market sizes,” says Bonte.

In that same vein, Andreas Mai, Cisco Systems’, director of smart connected vehicles, says that smart vehicles alone aren’t enough. “You need to connect things. You need to connect vehicles with traffic lights, with parking structures, with signs – and that is just the beginning. You need to connect data – pools of knowledge. For example, driving on a road, one needs to understand how good the drivers are that drive in those cars. You also need to understand how dangerous this particular stretch of road is, and you need to understand how healthy the operation of this particular vehicle is. By combining these pools of knowledge, you can identify many things, such as if there is a high or low risk of an accident.”

The Reality Check
To make fully autonomous vehicles a reality will require a number of elements be in place. First and foremost is safety. Second is ubiquitous connectivity – everywhere. That means redundancy, resilience, and availability with a five-nines guaranteed uptime.

Much of this will be accomplished by the integration of the next generation of Advanced Driver Assistance Systems (ADAS). This technology integrates preemptive instead of reactive models – to prevent an accident from happening in the first place, in some instances by taking control of the car away from the driver.

“Current ADAS systems are already pretty sophisticated,” says Freeman. “Cars with ADAS systems are able to brake if they get too close to the car in front of them, as well keep the car in the correct lane if they start to drift. These systems are still pretty expensive and limited mostly to high-end cars. With the technology advancements expected in the next 5 to 10 years, you will see that technology become more mainstream and find its way into mid- and low-end vehicles. At that point, everyone will have some of these ADAS feature for safety.”

Next is authentication and security. The network and the vehicle must establish a trusted relationship via encryption algorithms, secure keys, and network-generated “challenges.” The ramifications of corruption of the network are huge, so the network must be as close to being hack proof as is humanly possible. The biggest threat to the smart car infrastructure is that future smart cars will be electric and umbilically tied to computers and other smart devices as well as the power grid. That not only presents a problem for the smart car, but any hacking can be vectored back to the grid and any smart devices connected to the car.

There are a couple of bleeding-edge technologies that will be seen in smart cars. The most secure will be Face Detection Systems (FDS), fingerprint sensors and voice recognition. Most of these are still on the drawing board, but with the integration of crypto processors to protect keys and critical data, together they will create a very high level of security for the smart car.

Under the Hood
Smart car technology, like the rest of the IoT, will be sensor-based. Tomorrow’s smart cars will have a digital platform. The term Internet of Digital (IoD) is emerging as the platform for digital sensors. “We are now evolving to adding digital sensors such as cameras, digital RADAR, SONAR, LIDAR, which make the car more aware about its environment,” says Bonte. “It gets ‘eyes and ears,’ so it knows what is going on around it and can take appropriate action, autonomously.”

But there are still limits. “Today’s systems are still line of sight,” says Freeman. “They are better than the human eye, but are still basically limited to what the eye can see, although better in terms of sensitivity.”

The next level will integrate vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I), vehicle-to-home (V2H), vehicle-to-Grid (V2G), and vehicle-to-Person (V2P) communications, all which will be inextricably woven into the IoT. “This level of technology will allow the car to communicate with other cars, and also with the infrastructure around it. That will enable the car to know what is going much farther away than just line-of-sight technologies,” notes Freeman.

Missive
Once the smart car becomes mainstream, and a fully driverless transportation infrastructure is in place, it will resemble nothing near what we have today. “Such an infrastructure will eliminate blind-side intersection collisions, cross traffic accidents, passing hazards, and much more,” Freeman predicts. He goes on to say that another benefit is regulated traffic, which will make getting from points A to B much more efficient, and safer.

In fact, he believes the real issue is not about technology. “It is driven more by legislative issues, insurance concerns, global platforms, and those kinds of things.”

But there are a lot of disparate issues that plague an autonomous transportation infrastructure. There are topics such as security, liability, and cross-platform interconnect that are just scratching the surface of what really needs to be done. Whether it will be 25 years, or less as some predict, or the next century as others suggest, it will truly be one of the greatest marvels of modern technology.