In-Car, AI-Powered Speech Recognition Systems Now a Battleground
Automakers are integrating voice recognition inside the car; drivers are loyal to brands that offer better experiences; and car companies see revenue opportunities ahead.
By John P. Desmond, Editor, AI in Business
The car is now a battleground for AI-powered voice assistants, with the competition ranging from behemoths like Amazon to startups vying for a piece of the connected car market.
The connected car market is projected to reach a $215 billion value by 2027, according to Verified Market Research, quoted in a blog post on the site of SoundHound, an audio and speech recognition company.
The top three trends cited in the post are:
Auto manufacturers are increasingly integrating voice;
Drivers are loyal to auto brands that build better experiences;
And car companies are planning for new revenue opportunities.
Some 90 percent of new vehicles globally are projected to have embedded voice assistants by 2028, according to an account in Automotive World.
A huge incumbent player is Cerence, which spun off from voice recognition company Nuance in October 2019. While building its leadership position in voice recognition technology, Nuance had built a large customer base in the auto industry.
Cerence employed 1,300 at the time of the spinoff, including 700 in research and development. The company reported its AI voice technology was running in over 280 million cars on the road globally, across more than 70 languages and serving automakers including Audi, BMW, Daimler, Ford, GM and Toyota, according to an October 2019 account in voicebot.ai.
The CEO at the time of the spinoff, Sanjay Dhawan, stated, “The ability to focus solely on the automotive market will be hugely beneficial in achieving the next level of technological innovation in this space.” One characteristic of the market is that voice technologies from different providers may be operating simultaneously in the car. Or, drivers can be presented with a choice of a single voice solution.
Stefan Ortmanns took over as CEO of Cerence in December 2021. He had worked at Nuance for many years, including as executive VP and generation manager of the Automotive Division. Dhawan resigned to become CEO at SymphonyAI, an enterprise AI firm. Dhawan led the spinoff of Cerence and is credited for driving the market capitalization of Cerence from $500 million in 2019 to more than $5 billion last year, according to an account in the Boston Business Journal.
Cerence Co-PIlot Advances In-Car AI Software; Can Order and Pay for Coffee
In January, the company announced Cerence Co-Pilot, AI software connected to the car’s sensors that adds “voice, gaze, gesture and touch input” from drivers to assess what is happening inside and around the car. The AI can anticipate what the driver might want to know, such as weather updates or proximity to a favored stop for coffee, which it could order and pay for from a mile away. Or, if it knows the driver has an appointment some distance away the next day, it might suggest a fuel stop.
“Today, with the introduction of Cerence Co-Pilot, we place Cerence firmly at the center of reinventing what it means to drive a modern car,” stated CEO Ortmanns, in an account from voicebot.ai. “Voice-powered interaction has become ubiquitous in our daily lives, but it’s time that we expand upon its capabilities and chart the road ahead – a fully multi-modal, multi-sensor, AI-based experience. By bringing a new level of intelligence to the voice assistant, we not only enhance comfort and convenience, but also improve safety through proactive and reactive capabilities.”
Earlier in 2021, Cerence had announced: Drive 2.0 and connected Cloud Services, to enable the AI to better understand multipart commands; Cerence Look, which combines info from gaze-tracking cameras to observe what a driver is looking at; and Cerence Tour Guide, to provide insight on the area around the car.
“In-car assistants will transform into intelligent, multimodal co-pilots able to anticipate what the driver needs and wants and – equally importantly – know when the time is appropriate to deliver them that information,” stated Sujal Shah, VP & GM, Professional Services for Cerence, writing on the company’s blog.
Alternatives for drivers for in-car speech recognition applications include Apple CarPlay and Google Android Auto. CarPlay offers a version of iOS to the car’s touch screen display, fully integrated with Siri. If the driver receives a text, he/she will see a notification on the CarPlay screen, and with a tap, Siri can read the message aloud, according to an account on the blog of Summa Linguae Technology, a language and technology service provider headquartered in Poland.
Android Auto connects via a USB cable and relies on Bluetooth for voice phone calls through the car. Over 500 models are supported. Android Auto displays information including calls, texts and maps. Google plans to supersede Android Auto with Google Assistant Driving Mode.
Read the source articles and information in a blog post on the site of SoundHound, in Automotive World, in voicebot.ai, in the Boston Business Journal, on the blog of Cerence and on the blog of Summa Linguae Technology.
(Email the Editor here.)