Artificial Intelligence, autonomous vehicles, Deep Learning – these aren’t just buzzwords. The automotive industry is rapidly changing and in large part, it’s due to user demand. In these smarter, safer cars, it’s your interactions that will implicitly train the machine to learn your likes and dislikes – something made possible through these four pillars of AI.
The automotive industry is rapidly changing. Autonomous and semi-autonomous driving are key trends everyone is talking about. They will not only change how we use cars, but—in the long term – how our (smart) cities are designed. Autonomous driving became possible by recent breakthroughs in Artificial Intelligence technologies—Deep Learning is one example.
There are other, very exciting ways Artificial Intelligence technologies help improve the driving experience. Case in point: Intelligent personal assistants. While most assistants proliferate on a smartphone, there is innovation happening behind the wheel that brings an automotive assistant to bear that fulfills your driving-related needs better, faster and safer. These are the four main pillars in AI being applied to the connected car:
We make it possible for people to talk to and interact with machines – an experience that continues to become easier over time with advancements in Natural Language Understanding . In fact, being able to speak to your devices and cars is an expectation. You can say what you want in almost any way you like and have longer conversations. Now, the machines will remember what you said before, even if it was yesterday, and use that to get to the desired outcome more quickly and with less effort. Also, just like humans, the car will be able to recognize you by your voice through voice biometrics technology, which is useful to distinguish you from your spouse, for example.. That means even more personalization. So when you use your spouse’s car and say “Let’s drive to work,” the car will get you to your workplace, not your spouse’s workplace.
Since the car, by definition, is moving from place to place, the contexts in which it operates change frequently. You might be on the highway, or on a dirt road in the countryside. Maybe there is a traffic jam ahead, or a road closure. There may be sunshine at your current location but a snowstorm at your destination. Each of these can provide additional information that will make your interaction with the car more meaningful.
Context also includes sensor reading from the car itself, e.g. fuel level, crash avoidance sensors, or how many people are in the car. These parameters matter for making smart decisions when driving, and are taken into account by the contextual reasoning framework . Essentially, it’s artificial intelligence that thinks for you. The system knows that in a snowstorm, covered parking is certainly the preferred option. It will also try to find the cheapest parking for you based on your estimated arrival time and will find a gas station on your route that requires only a small detour, but is cheap at the same time. If you have a loyalty card, that will also be considered. It will even recognize that the traffic jam ahead will likely cause you to be late for your meeting, and offer to send your colleagues a note about your delay. These ideas aren’t simply flashy; our research shows that users are excited about cars that can make this a reality.
Until fully automated vehicles arrive, decisions while driving can only demand a certain amount of attention from drivers, since the primary focus should always be on the road to enable a safe journey. But in many situations this is difficult to realize. Say you are looking for a parking space, or a restaurant for dinner, or you want to create a playlist with a certain type of music. In all of these situations there are too many options to choose from to make an informed decision while staying focused on the road. Enter personalization.