Automotive HMI design: How AI can save the identity of car brands

With the ever-increasing connectivity of the modern car, there is now more information and more functionality available than ever before, shifting the design focus to Human Machine Interfaces (HMI) rather than the look of the car. This poses a complex situation for car manufacturers as they work to determine how AI and smart arbitration will impact the future of their brands.
The ever-increasing connectivity of the modern cars Artificial Intelligence is shifting the design focus to Human Machine Interfaces (HMI) rather than the bodywork

Distinct design has always been at the heart of the car’s appeal.

Buying a car is not a rational undertaking – the emotional pull of a car brand is often a key factor in a customer’s buying decision. So, it’s natural that car manufacturers have always spent a lot of effort on creating a distinct design and user experience (UX) that separates their cars from others. And they have largely succeeded: car brands transport certain images, and they have done so for a long time. Some models have even become iconic staples of popular culture; just think of the VW Beetle as the symbol of the German “Wirtschaftswunder” of the 1950s, or the irresistible coolness of the Ford Mustang, racing through the streets of San Francisco with Steve McQueen on the steering-wheel and encapsulating the spirit and style of the 1960s.

But with the ever-increasing connectivity of the modern car, there is now more information and more functionality available than ever before, shifting the design focus to the Human Machine Interface (HMI) rather than the look of the car. Virtual assistants are now commonplace in modern cars, and their design is paramount to the driver’s identification with the machine. It’s not only how your car looks, but also how well you can interact with it.


New competition emerges.

With connectivity deeply changing the industry, new players like Google and Apple are strongly pushing into the automotive market with platforms like Android Auto and Apple CarPlay. These platforms seem to have one major advantage: a seamlessly integrated UX. Players like Google and Apple know that the mobile handset is the main hub of connected users. We organize our schedules, stream media, plan trips, and stay in contact with others, all on our phones – and we expect it to be available when we enter our cars. With this in mind, the move into the car seems natural for those major tech companies.

This poses a complex situation for car manufacturers. If they refuse to strategically engage with the expectations of connected users, they will lose customers. But, if they let other companies take over their HMI, they are sacrificing an important part of their design. Their cars in effect become empty vessels for another company’s design, and with that they lose an important argument: why buy this specific car when all cars are powered by the same handset centered, general purpose assistant?


So, are the design aspirations of manufacturers a lost cause? Not quite.

There are three key factors in favor of automotive manufacturers and their suppliers. First, they have vast experience in the automotive business. What sounds trivial is actually crucial. Their experience with drivers’ needs allows them to offer solutions that are tailormade for the automotive use case and offer better functionality than general purpose solutions. Just imagine how many parameters you can use to look for a parking spot. Is parking free or is it paid? How much does it cost? Do I pay with cash or with credit card or with electronic payment? Is it secured? What are the opening hours and will it still be open when I get there?

Second, car manufacturers can offer embedded solutions with crucial advantages. An embedded solution in a modern car offers deep integration of the HMI with the manifold sensors, far beyond any level of integration a mobile solution can offer. As we witness the dawn of autonomous driving, deep integration and the ability to access and leverage the sensors without gaps or delays will become crucial. If you hand the steering wheel over to a virtual assistant, you want it to correctly understand all the information the car offers. That is much more likely to be achieved by an embedded solution that is tailor made to this car’s specifications. Furthermore, the embedded solution is not dependent upon unrestricted access to the cloud. If you take your car outside of urban centers or into a parking garage, online connection is still hit or miss. And you certainly don’t want your navigation guidance failing while you are in the middle of nowhere.

Third, manufacturers can integrate the functionalities of multiple virtual assistants while maintaining the priority of their own system and design. To do this, they need an automotive assistant with cognitive arbitration capabilities that knows when a request needs to be passed on to another assistant. For example, if you are driving home on a cold winter night, your smart assistant can contact your smart home assistant and heat up the living room or order a pizza. The automotive assistant does not need to do everything – it just needs to be able to distribute tasks to the most appropriate agent. That level of smart arbitration is made possible by the latest advances in artificial intelligence. Features like voice biometrics allow assistants to easily identify different users by the sound of their voice and load their personal preferences. The assistant can also continue to learn. If I order a pizza, the assistant should know that I like my pizza with olives while my wife prefers pepperoni. How does it know this? Because that’s how it was ordered the last time. Context awareness allows us to further refine arbitration: if the system knows I am driving home, it makes sense to contact the home assistant and have it order a pizza for me. If I am driving around in another city, it might make more sense to use the navigation system to help me find an Italian restaurant.


Challenges and opportunities ahead.

A car with a unique design is still a thing of passion. At the same time, connected users expect unrestricted and seamless functionality that extends into their cars, and Silicon Valley is expanding into the automotive market to meet that need. At first glance, what seems like a problem for traditional car manufacturers could turn out to be their biggest asset. They need to reconcile the importance of distinct design with the need for connectivity over multiple devices. If they can design a smart assistant that offers a unique voice and knows who is best at doing what, then they will be in a very good position to maintain their status as design innovators.

More brains are better than one:

Get to know our AI-based cognitive arbitrator that connects virtual assistants in the smart home, car, and enterprise.

Learn more

Tags: , , , ,

Tarek Sumiri

About Tarek Sumiri

Tarek Sumiri is a Sr. User Interface Designer in the Nuance Automotive Professional Services team. As such, he is constantly involved in developing state of the art speech systems for automotive customers. As head of the localization team, he is bringing applications to markets worldwide. Tarek holds a Master’s degree in Political Science and Sociology and also studied Philosophy and Communication Studies. He likes to take a step back to see the bigger picture and to reflect on human-machine relations, on the societal impact of technological change and on the power of aesthetics.