The car of the future will be as attuned to our needs as a personal assistant. It will know our wishes, entertain us, even speak with us. The experts at Volkswagen Electronics Development are working hard to make this goal a not-too-distant reality. Part 1 of our series on user experience trends.
It’s a bitterly cold winter day as Stephanie packs her bag for a trip to the gym. A slight vibration tells her to pick up her smartphone, where she reads a new message: “Hello Stephanie! You’re probably just about to head off. Should I warm up for you?” Practically no one knows the 40-year-old as well as her car. “Yes, please warm up,” confirms Stephanie. In 20 minutes, she will set off for the gym – and start her drive at a comfortable 21 degrees.
What sounds like a pipe dream could be commonplace for many drivers in just a few years. In Volkswagen’s Electronics Development department, experts like Kathrin Wilkens, Manuel Joachim and Kord Lühr are already working on enabling cars to intuit drivers’ wishes even before the trip and adjust to them optimally. They created the fictitious gym-goer Stephanie for their development work in order to get as realistic an idea as possible of her needs. Thanks to new technology, there may really be customers like Stephanie in the near future.
The right temperature at the right time
Kord Lühr is the development team’s expert for the right temperature at the right time. “We teach the car to understand the driver’s habits better and better over time,” he says. The idea: Almost every person follows recurring patterns and habits. Just like Stephanie. Every morning at 7:30 a.m., the comptroller drives to work, and she usually heads home between 4:00 and 5:00 p.m. Almost every Monday at 6:00 p.m., she goes to the gym.
“Artificial intelligence can recognize such processes as patterns and respond accordingly,” explains Lühr. The advantage: If it’s cold outside, the vehicle of the future could remind Stephanie to switch on the engine-independent heating – something she might otherwise forget in the hectic to-and-fro of everyday life. “Artificial intelligence is making enormous strides. Soon it will be able to understand us so well that the car will go from being a mere form of transportation to a personal assistant as well,” says Lühr. What’s important to the developers is that the user remain in control of their data at all times. Anything that he or she doesn’t want to reveal will not be evaluated by the assistant.
Fast facial recognition is the key
As soon as Stephanie is in the vehicle, she can access her email and social media accounts, payment options, or the functions of Volkswagen We. Her seat, exterior mirror and infotainment are also tailored to her. But how does the car know that Stephanie is really Stephanie? Precisely that question is the focus of Manuel Joachim, the development team’s expert for biometrics. “Our goal is for the car to recognize the driver as they’re getting into the vehicle,” he says.
To achieve this goal, Joachim and his colleagues are testing identification methods using various biometric characteristics such as fingerprints and irises. The expert considers facial recognition via infrared camera to be especially promising. “The advantage is that the user doesn’t have to do anything more than simply look through the windshield. Everyone does that as soon as they sit in the car anyway.” Unlike with fingerprints, drivers would not even notice that the vehicle was making a fast scan of their face. They would simply get in and presto, all of the usual functions would be immediately available to them – exclusively.
The best routes are already calculated
Stephanie is sitting in the car. The temperature is pleasant, and the driver’s seat has automatically assumed the optimal position. Stephanie looks at the display, where the car is suggesting the best routes to her probable destinations. With a single command she can start the navigation to the desired one. How this functionality will work in the future is the field of developer Kathrin Wilkens. She is preparing artificial intelligence for the task of calculating Stephanie’s probable plans for the evening based on a variety of data sources. “If Stephanie wants it to, her personal assistant will understand her behavior better and better over time,” says Wilkens. It will, for example, be able to predict that Stephanie will want to go to the gym on Monday evening with a probability of 78% – compared to 10% for the supermarket and 8% for the movies.
The more information that Stephanie shares with the assistant, the better the predictions will get over time. “One of the most important data bases for good predictions is previous drives,” explains Wilkens. But also entries in the personal calendar or an order made at an online retailer can provide valuable information about the person in question. “Where Stephanie really wants to go will, of course, be known by her alone even in the future,” says Wilkens. It might be that she recently bought concert tickets for the following Monday. “But maybe she gave the tickets to a friend and she herself is going to the gym as usual. The car can’t know that, of course.
And what if Stephanie is driving not in the winter, but in the sultry height of summer? Even at very high temperatures, her future drives can also be started much more comfortably than is the case today. “We’re working on enabling the car to recognize users as soon as they approach the vehicle,” says developer Kord Lühr. Rather than warming up the car, the personal assistant could simply turn on the ventilation. The advantage: In contrast to heating, even brief ventilation can make the temperature in the car much more pleasant. For Stephanie, this means she can start sweating at the gym rather than on the way there.
Working in Electronics Development
Electronics Development at Volkswagen has opportunities mainly for electrical engineers, computer scientists, natural scientists, psychologists (human-machine interface), mathematicians and software developers.
Hard skills in demand
- IT know-how: networking technology, server technology, object-oriented programming languages such as C++, JAVA, software development for embedded systems, ideally in the automotive field
- Knowledge of tool-aided modeling of software architectures in UML as well as the use of design patterns
- English-language skills
Soft skills in demand
- Ability to communicate / Team skills
- Positive approach to flexible working conditions
- High degree of personal initiative
- High analytical competence