Skip to content

Getting louder – the rise of the voice in automotive HMI

Major advances in natural language and voice recognition mean that talking to your car will soon be the norm, writes Rachel Boagey

Keeping a driver’s eyes focused on the road ahead has always been of great importance, but with the increase of in-vehicle connectivity demands, OEMs are under even more pressure to reduce driver distraction. Many OEMs such as Tesla are already scrapping the use of the buttons in their vehicles, favouring touch screens and speech-interfaces, which now need to be more efficient and user-friendly.

The use of voice applications can help to bypass numerous menu stages, and control many in-car functions thought to be distracting to the driver. Despite this, many argue that voice is not suitable for controlling all in-car applications, often taking longer to perform complex tasks than manual administration. With that in mind, is voice the logical next step towards safe in-vehicle interaction?

Driver relevant content

Arnd Weil, Vice President and General Manager of Automotive and Consumer Electronics at Nuance, recently spoke to Megatrends regarding the role of voice in the automotive HMI, saying, “Consumers want access to the apps and content they know and love on their smartphones, coupled with apps specific for the in-car experience.” However, Weil suggested that “these experiences must be accessed in a way that avoids dangerous driver distraction.” But to achieve these goals, there are many hurdles for OEMs to jump, not least of which are the importance of delivering a great user experience, and providing more than what already exists in mobile handsets.

Nuance is a supplier of speech solutions for OEMs such as Audi, BMW, Ford and Hyundai, and has developed Dragon Drive Connect, an automotive grade in-car service that integrates content services and applications that have been optimised for in-car use. Using Dragon Drive Connect, drivers can stay focused on traffic and driving conditions by interacting with the in-car systems in a hands-free and eyes-free way. “Voice recognition, natural language understanding and text-to-speech technologies enable intuitive interaction with the car, thus reducing possible driver distraction,” explained Weil.

Drivers don’t need access to thousands of services while driving, he explained, maintaining that this would in fact serve as more as a hindrance than a benefit to safe driving. “Drivers need access to content and services that relate to the driving experience, such as services focused around navigation, telematics, infotainment, communications, and vehicle maintenance.”

OEM developments

Ford puts its efforts into hands-free control and minimising driver distraction, and its Ford SYNC in-car system allows users to control a wide range of in-vehicle systems using their voice. Ford’s SYNC uses speech technology by Nuance. Fatima Vital, Director of Marketing, Automotive & Consumer Electronics at Nuance, told Megatrends, “Voice recognition, natural language understanding and text-to-speech technologies enable intuitive interactions with the car, thus reducing possible driver distraction. We expect that voice interactions will become the main way of accessing content and apps in the connected car – closely integrated with other modalities, such as touch, handwriting or gesture control.”

Julius Marchwicki, Ford SYNC Product Manager, explained to Megatrends how the OEM makes safety a focus of its HMI: “Complex tasks should be able to be executed with a simple voice command within the car, and most OEMs now are strongly leveraging voice. The technology is only getting better and creating better experiences for consumers inside the car. One of the first things we think about is safety and we are looking to provide value-added services for the consumer. With the additions of new capabilities and functionalities, we always look at driver distraction testing and safety testing to try to get the best experience possible.”

Jim Buckzkowski is the Global Director of Electrical and Electronics Systems and Advanced Research at Ford, and spoke to Megatrends about how effective voice applications are for tackling driver distraction. He said, “We do a lot of research here on driver distraction and have a simulator that allows us to test people in a safe environment but actually in a car inside a chamber where we can simulate actual driving. We focus on how long it takes for drivers to complete a task, and monitoring the driver’s gaze to see how often they’re looking away and for how long. We know it’s really important that looking away from the road should be short and not sustained for a long period of time.”

Buckzkowski also explained the importance of OEMs not allowing drivers to be faced with tasks that require repeated glances away from the road. He said, “We spend a lot of time looking at different ways of designing human machine interfaces, and test these different interfaces and different HMIs and so on to see what works. Voice recognition definitely is a significant improvement over picking up and hand-holding a device and trying to operate that device in your hand while you’re driving.”

Voicing an opposing view

Despite its benefits to driver safety and its ability, unlike other applications, to minimise driver distraction, Bryan Reimer, Research Scientist at MIT’s Age Lab believes that although there is a great role for voice in cars, its benefits aren’t always as good as they seem.

To reach this conclusion, MIT carried out a study with support from Toyota’s Collaborative Safety Research Centre, finding that using voice commands for simple tasks, like searching for a radio station or changing the climate settings, is quicker and less taxing for a driver. But using voice controls to perform more complex tasks—such as tuning the radio—often takes longer than doing those things manually, causing drivers to glance away from the road to select from a menu or confirm that the system has recognised their speech correctly.

Reimer told Megatrends, “Any non-driving related activity can become a safety relevant distraction if a driver is engaging with it at an inopportune time. Our work as well as the work of others shows that “voice command interfaces” are not hands free eyes free ways of communicating with your vehicle. They draw upon visual, manipulative and cognitive resources that must be measured and considered in a holistic view of resource allocation.”

Driver choice

Luc van Tichelen, VP, Embedded Professional Services – Automotive at Nuance, explained that although voice applications are useful in the car environment, certain tasks can’t be done at the wheel without distracting the driver. He said, “Some things aren’t as important while driving, so it is essential that notifications such as a new email isn’t an attention-getter in the car.” In this case, explained Tichelen, “The driver should be allowed to pick up the information when they are ready to look at that screen, in the form of a small pop-up rather than a voice notification.”

Unsuspecting driving aids

Waze is a smartphone app that drivers will need to bring into the car. However, the company’s Head of Global Communications & Creative Strategy, Julie Mossler explained the safety benefits of the app, which provides turn-by-turn navigation and direction via audio, meaning the driver does not have to look at their phone. “Safe driving is core to everything that we do,” said Mossler, “and all of the features that we choose have to pass the safe driving test. As a global driving society, GPS is largely excessive in the car, so much like a GPS, we recommend that docking your phone in a proper car dock and plugging it into a power outlet is going to give you the best results.”

Lost in translation

Dealing with language barriers and accents often proves difficult for voice control and recognition, but Mossler explained that Waze’s smartphone navigation app is available in 45 languages and 200 countries. She said, “I think that you eliminate a lot of that risk when users can speak in their native language. And, hopefully, if you speak one of the 45 most common languages in mobile technology you’re going to be just fine. Aside from that, we also do have manually written directions that you can look at on the screen. So, between those two things we don’t really hear about language being an issue.”

Ford’s Buckzkowski also explained how the OEM’s HMI caters for different regions and languages, saying, “Voice recognition technology continues to get better at adapting to more extensive languages, while also becoming more conversational, rather than just providing words and commands.”

He continued, “HMI is placed on the screen differently based on languages in different regions of the world. We also know that in some regions people are very quick to adopt to talking to their car. Others are maybe a little bit more shy about it. But there is a possibility that it’s even more generational than it is cultural.”

Harsh vehicle environment

The challenging audio environment of a car can also be a barrier to effective voice applications, and noise that comes from the car, from road noise to crying children, can be factors that judge how reliable voice applications really are. Developments in hardware, software and connectivity are required before voice technology can be really successful, especially when coping with difficult instructions.

Danny Shapiro, Director Automotive, NVIDIA, explained to Megatrends, “I think there are two aspects of making sure voice applications are effective in the car: one is making sure you have good microphones, and multiple microphones in the car, and two is software; a lot of software is needed to be able to enable noise or echo cancellation. There’s a huge amount of research that’s been going on in that space, but now we need to bring it into the vehicle.”

Creating a voice for the older generation

A consideration for OEMs to bear in mind when implementing their infotainment systems into cars is the specific age categories for which they are catering. As the younger generation may want complex technology in their car, it becomes even more important to use voice technology to ensure they are not getting distracted from the main task at hand. However, another consideration is older drivers. The general consensus is that voice recognition will be used to perform more complex tasks in a vehicle as it can help bypass many menu stages that previously would have left drivers’ eyes off the road for long periods of time.

Nuance’s van Tichelen also explained the importance of considering the older generation of drivers, saying, “These drivers will prefer using the information from a functional point of view. They don’t necessarily need all of the advanced features, and often hate looking for what they need.”

An audible future?

Despite developing rapidly, voice technology remains a contentious issue. Ford’s Buckzkowski told Megatrends, “We know there is room for improvement, but it’s going to continue to advance in the future.” Buckzkowski is one of many who think voice will become prevalent in vehicles going forward, with developments of voice applications in the automotive industry so far showing the potential for a successful future. However, it is clear that much more work is needed to make voice applications largely successful.

This article was first published in the Q2 2014 issue of Automotive Megatrends Magazine. Follow this link to download the full issue

https://www.automotiveworld.com/articles/getting-louder-rise-voice-automotive-hmi/

Welcome back , to continue browsing the site, please click here