Is AI in Cars the Solution to Driving Safer, More Efficiently, and Happier?
Exploring the challenges of voice assistant technology in vehicles and the need for advanced AI solutions to improve safety and accuracy in diverse driving conditions.
AI's Role in the Automotive Industry
Artificial intelligence (AI) was a significant presence at CES 2025, especially in the automotive and transportation sectors. In his keynote address, NVIDIA CEO Jensen Huang underscored AI's potential, emphasizing how AI and advanced computing will reshape how we think about and design vehicles.
This influx of AI-powered automotive technology raises important questions. Can these innovations truly enhance driver safety, streamline the user experience, and optimize travel efficiency, or are they merely flashy gimmicks? Can AI genuinely address the current challenges drivers–and OEMs— face by achieving accurate voice recognition against background noise, simultaneous speech, and correctly locating the desired source speech, or is it all style over substance, as some industry experts suggest?
A Personal Anecdote
I used to drive 76 miles from Santa Cruz to San Francisco, which took two to three hours. The long drives were exhausting, resulting in fatigue, speeding tickets, and getting lost. Today, technology has not improved my experience on this drive.
My car’s infotainment system connects to my iPhone for voice navigation, but my car's assistant (seven years old) is outdated and ineffective. Siri, my iPhone’s voice assistant, often can’t hear me, misunderstands me, or provides incorrect results, making me want to throw my phone out the window. Siri struggles to hear me because of my distance from the microphone and the road noise. The language model may also pose challenges, requiring me to manually operate the infotainment system, which is unsafe. Google Maps helps with navigation most of the time—when Siri understands my directions.
Poor Voice AI Performance Not a Unique Problem
It’s not just me. The problems with voice technology in cars have long been studied and documented. A recent article from The Autopian highlights that these systems often suffer from slow response times, misunderstood commands, and limited functionality, making them less efficient than traditional controls.
Considering these frustrations with current technology makes me curious whether the AI innovations showcased at CES could address some of the issues I and many drivers encounter. Maybe.
Promising AI-Driven Technologies
Some of the AI-driven technology I saw at CES seemed fanciful and distracting. Others seemed like they could be helpful, such as Bosch’s wrong-way driver alerts system or Seeing Machines’ driver monitoring technology that monitors driver attention to reduce fatigue-related incidents. Additionally, BMW's Panoramic iDrive and Hyundai Mobis' Holographic Windshield Display intend to provide information to drivers clearly and safely.
Insurance Institute for Highway Safety researchers say the key to driver safety is limiting visual display distractions. However, Senior Research Scientist Ian Reagan told AP News he did see promise in keeping critical information in the line of sight; for example, not having to look down or over at an infotainment screen for directions. That’s what BMW’s Head of UI/UX Development Stephan Durach told AP News. The goal of the company’s Panoramic iDrive was to keep the most critical information projected onto the lower windshield of its cars to minimize distractions.
Challenges with AI and Voice Assistance
However, the lingering question is whether AI—utilized for navigation, entertainment, informative displays on windshields or dashboards, or to keep me alert—will enable my voice assistant (whether on my phone or integrated into the car) to hear, comprehend, and respond accurately to my requests.
This technology should allow me to do so without requiring me to repeat myself, shout, glance at my phone, or interact with the car’s interface, all of which pose dangerous distractions while driving.
Based on what my colleagues and I experienced at CES, the answer would be no.
One of our engineers attended a demo of a popular voice assistant in a luxury car brand. During the demo, the assistant answered questions and personalized responses in a quiet cabin. However, it failed to understand the passengers in the back seats. I question whether it could have understood those in the front if there had been background noise and crosstalk. The experience is not unique.
Solving the Problem
OEMs and Tier 1 suppliers must prioritize integrating advanced AI technologies capable of understanding speech in diverse environments to address poor voice recognition and inaccurate results from car voice assistants. This includes enhancing the clarity and precision of voice recognition systems to accurately interpret commands, even with background noise and multiple speakers present.
Collaborating with leading AI firms, like Kardome, can help OEMs develop robust solutions that can isolate and comprehend user voice inputs from various cabin locations.
Additionally, ongoing user testing and feedback loops should be established to continually refine these systems, ensuring they evolve alongside real-world driving conditions. OEMs can significantly enhance consumers’ overall driving experiences and safety by focusing on reliability and user-centric design.
A Solution: Spatial Hearing AI Technology
Fortunately, there's a way to address the issues mentioned—Kardome’s Spatial Hearing AI technology. This innovative voice AI solution can locate, isolate, and comprehend what someone is saying, regardless of whether they're seated in the front or three rows back. It can do this even amid music and overlapping conversations, as showcased during our in-car demonstrations at this year’s CES.
The evidence speaks for itself: Kardome’s solution will be integrated into millions of vehicles this year by one of the largest automotive manufacturers in Asia, combined with existing ASR technology from another reputable company that lacks Kardome's capabilities.
A Final Word
While the automotive industry is abuzz with AI-powered advancements, their success is in their ability to address the daily challenges drivers face. While some technologies show promise in enhancing safety and improving the driving experience, the persistent frustrations with voice recognition systems highlight the need for more sophisticated solutions.
Kardome's Spatial Hearing AI technology offers a breakthrough, demonstrating that it's not just about adding flashy features but fundamentally transforming how we interact with our vehicles. As AI continues to evolve, the hope is for a future where technology truly complements the driving experience, making it safer, more efficient, and, yes, even a little bit happier.
