IoT Blog
IoT Blog

Self-Driving Cars of the Future: Reliable IoT Components Promote Vehicle Safety

by Andreas Kohn, Director of Marketing, Automotive Solutions
Driverless cars have long been a favorite among science fiction writers, futurists and countless Hollywood films. But now, we’re on the cusp of seeing autonomous cars shift out of the realm of imagination and into reality. When trains were first introduced as a reliable mode of transport, some scientists believed that moving so quickly would prove invariably fatal for women. When cars were invented, there was a worry that driving faster than 60 would be fatal for anyone—a more egalitarian paranoia. Even today, with air travel considered fully mainstream, a full 25% of the population is nervous about flying and about 6.5% suffer from clinical aviophobia

But the concerns surrounding self-driving cars are unique. It feels instinctively and viscerally unsafe to take your hands off the wheel. Self-driving cars will, of course, be considerably safer than human-operated vehicles. In fact, in the U.S. alone, it’s estimated that a system of driverless cars could reduce fatalities by up to 90%, while saving $160 billion a year in gas and lost productivity from traffic jams. OEMs are now tasked with helping the public overcome their fears about autonomous vehicles. 

blog cta - open source

To convince consumers that self-driving cars are truly safe, the industry must be unwavering in its emphasis on consistency, reliability and intelligent engineering. The benefits must far exceed the shallow novelties, such as, “You can check Facebook during your commute!” Driverless cars must be proven to be safe. They must be beyond reproach. One way to achieve this is to utilize reliable IoT components that maintain a stable connection and allow for easy updates. 

The right connectivity technology will propel self-driving vehicles into the mainstream, successfully overcoming the political, logistical, cultural and psychological barriers that exist today.

What Do Connected Cars Need to “Talk” About? 

Many drivers don’t fully appreciate the connectivity of today’s cars. In fact, nearly every new car on the road has some level of connectivity. By 2020, it’s estimated that 30% of all new cars worldwide will have embedded connectivity—up to 75% in the US and Europe—an increase from 13% in 2015. Right now, that manifests itself mostly in infotainment and analytics. Data collection—used to understand what happens on the road, including every pothole, every sudden stop, every fearful blind alley—is a major part of connectivity. Connected truck fleets have been using this technology for quite some time. This gradual shift toward autonomy and vehicular intelligence is a natural lead-up to the self-driving car of the future.

Data collection is a key element of the equation because as cars see greater levels of autonomy, the vehicle requires a greater sense of self-awareness and awareness of the world that surrounds it. Developers are in the midst of creating sophisticated algorithms that are both adaptable and predictive. Therefore, if the self-driving car “sees” a pedestrian step into the street, it will have the ability to predict the pedestrian’s actions (i.e. crossing the street) and the vehicle will instantly know what to do. These highly sophisticated vehicles will hold the ability to “learn” over time, as the car stows dozens of similar situations in its memory banks. In time, this data will simulate human instinct, allowing for prediction and reaction. 

But self-driving cars need more than the ability to sense surroundings, predict behavior and react in real-life scenarios. They must communicate with one another and with the surrounding infrastructure. The industry has developed a concept known as V2X Communication: that is, vehicle-to-everything communication. This all comes in addition to the typical connectivity capabilities that will allow riders to browse the web, take part in virtual meetings, stream movies, and listen to music during what promises to be a stress-free commute. 

High-tech vehicles will be at the center of what will be a constant and ever-growing data stream, flowing from the millions of vehicles on the road. But before this happens, we need the right connectivity. As cars become more autonomous, their mission-critical functions will require a connection with unparalleled reliability. If that connection is interrupted, the consequences could be potentially disastrous. An entire city’s worth of roadways—designed to run in graceful algorithmic elegance—could come to a grinding and implacable halt. In today’s time-pressed society, that just isn’t an option.

The 5G Connectivity Model
So how can this stoppage be avoided? How do we avert a disaster that could prompt people to question the safety and viability of self-driving cars? After all, if a car can stop when it loses connection, there’s a possibility that it could swerve off the road. Cars could collide with each other, or worse, pedestrians. A single hacking incident could hold the power to cause mass casualty. While these scenarios are, at best, far-fetched (if not outright fictional), the fear is very understandable and real. 

The key is to use the right connective tools; tools that can handle enormous volumes of data. There are still some questions as to what this will look like in practice. Since there are different kinds of data being transmitted, there might be different connectivity modules. In theory, you could have two or three modules placed in every car. The first would be for mission-critical operations, such as driving (here, “mission-critical” seems like an understatement). Manufacturers could integrate additional modules for “nice to have” components, such as infotainment or usage-based insurance monitors. 

Future 5G networks should open this new realm of possibilities. 5G networks are slated to come online by 2020, so it’s possible that autonomous vehicles could take to the roads shortly thereafter. While Dedicated Short Range Communication (DSRC) tools are good for vehicle-to-vehicle connections, they are generally viewed as old technology that’s simply not capable of handling the vast quantities of data. New standards must be implemented. 

3GPP, the global technology standards body, has standardized in Release 14 a physical layer (V-LTE or “Cellular Vehicle-to-Everything [V2X] standard”) based on LTE. Optimizations will come in Release 15 and 5G. This wireless technology needs to be implemented in cars in order to connect with the networks that will ensure consistent communication. 

That means in-car connectivity modules must be reliable, whether they are for mission-critical or second-level operations. Internet of Things-enabled modules have established a proven track record for efficacy, reliability, and security, and durability standards across 2G, 3G, 4G, and LTE networks. Most agree that these IoT modules will serve as the most effective option for ensuring a smooth transition from driver-operated cars to autonomous vehicles.

There are still a few technological and social hurdles to clear before we have a connected network of self-driving cars that will make life easier and safer for everyone. The great thing is that overcoming the technological challenges, the primary one being reliable connectivity to increase safety, makes it easier to overcome the social ones. Start with Sierra to learn more about how driverless cars are nearly a reality.