Time to Get Personal

Industry engineers are combining apps, voice, the cloud, and other technologies such as artificial intelligence to enable drivers to customize their vehicles and anticipate their needs.

January 1, 2016

Terry Costlow

Artificial intelligence will be used as Visteon creates links between cell phones and dashboard displays.

The race to let drivers personalize vehicles is gaining speed. A range of techniques and technologies are being deployed to let drivers do more to make vehicles as individualized as their consumer environment.

Design teams are pushing personalization well beyond simple tasks like moving seats and picking radio stations to match instructions in key fobs. Apps and other techniques that personalize rental and ride-sharing vehicles are emerging, as are technologies that alter settings even when multiple people share a key fob.

“We work with a speech-recognition company that can tell the difference when you say ‘hello, car’ and when I say ‘hello, car.’ It can also prioritize when more than one known person is in the car,” said Andrew Poliak, Global Director, Business Development for QNX Software Systems.

Minimizing inputs to reduce driver distraction is a central part of Aupeo’s strategies.

Clouds and context

The scope of attributes that can be adjusted is expanding rapidly. A growing number of companies are using the infotainment system to combine context, such as vehicle sensor information, and the content, that includes information available through cloud services.

MirrorLink can help drivers manage smartphone apps using vehicle controls.

“The infotainment system’s new role is to combine context and content to create value for the end user,” said Olivier Charra, Senior Engineering Software Architect, Automotive Solutions, at Wind River. “We are moving away from an app-centric model to a user-centric model, with the ultimate goal being to actively present to the user the information he needs, at the right time, with the right level of abstraction.”

Ultimately, vehicles should be able to learn parameters such as destinations. Simply altering routes as traffic changes won’t be enough. For example, interconnected systems can determine that a typically solo driver needs to navigate to school before going to work.

“If seat sensors show you’ve got kids in the car, it will switch the route,” said David Taylor, CEO for Aupeo.

Learning what’s normal

QNX is using voice controls to activate profiles that create the vehicle atmosphere users want.

This requires a lot of software. Many design teams are creating programs that learn what drivers normally do and how normality changes so they know what might come next.

“Artificial intelligence could be lever-aged to predict things such as radio and music preferences, destinations, routes, traffic alerts, reminders, and which contacts you might be likeliest to attempt to communicate with at any given time, just to name a few,” said TC Wingrove, Senior Manager for Product Innovation at Visteon. “Whereas smart-key personalization would be limited to a small number of keys for a particular vehicle, app-based personalization could theoretically allow personalization on any vehicle that can interface with the app.”

OEMs and Tier 1s who went through the early days of voice recognition may tread slowly before rolling out these advanced capabilities. Users expect high quality from vehicles. If systems mistakenly alter parameters, consumer satisfaction rankings may suffer declines similar to those seen when voice recognition failed.

“When vehicles do personalization, getting things wrong 10% of the time will be very annoying,” Taylor said. “The 90% of the time it’s right doesn’t seem like that big a deal.”

Vehicle systems will acquire some of the contextual information from the user ID. Each person’s profile contains information such as subscription services related to music, navigation, traffic, or weather. Connectivity makes it possible for powerful servers to analyze data from vehicles, performing assessments that are more complex than on-vehicle systems can perform.

“Once the car becomes connected, it can basically interact with any service exposing an application programming interface in the cloud based on events or on information from the vehicle itself,” Charra said. “The same way we transitioned from feature phones to smartphones, we’ll transition from feature cars to smart cars.”

Wait, don’t tell me

Software will be a primary element in personalization technologies. Source: QNX

The industry is generally focused on offering users more options, but having too much information and too many options can be confusing for people controlling a vehicle. Aupeo’s Taylor noted that in cars, less can be more.

“We’re moving to a level of personalization intended to help by removing unwanted information from the driver,” he said. “If there are no delays on the route to work, there’s no need to provide a traffic report.”

The push to personalize vehicles with apps extends the industry’s climb into the cloud. Most smartphone apps are basically user interfaces for cloud-based services. Vehicle systems may also use the power of remote servers.

“The actual service and its data are in a backend server in the cloud, with this service being in charge of the synchronization between all user interfaces,” Charra said. “The part of the application actually running on the smartphone itself is fairly limited. That won’t be different in the case of an infotainment system. The infotainment system will be just another interface to some cloud-based services.”

While the cloud may eventually handle many facets of vehicle customization, radio head units will still need a fair amount of computing power. Cellular links haven’t yet achieved automotive grade reliability.

“In an ideal world, most decisions would be made in the cloud,” Taylor said. “Unfortunately, connections aren’t always good enough, so data may only move between the car and cloud on intervals. Systems do need to be sensitive about the amount of data being sent, which does cost money.”