Artificial Intelligence & Machine Learning
How Generative AI Will Shape the Future of Auto Industry
Technology Is Disrupting the Way Vehicles are Designed, Manufactured and ExperiencedMercedes-Benz cars just got smarter. Powered by OpenAI's ChatGPT, its new voice feature "Hey Mercedes is rolled out as an optional beta programme in the United States for vehicles equipped with the MBUX infotainment system. The automaker is transcending conventional boundaries of AI applications and integrating heightened intuitiveness into its MBUX infotainment system. The company is integrating ChatGPT through Microsoft Azure OpenAI service, leveraging capabilities of Microsoft's cloud and AI platforms. Using this, users will be able to engage the chatbot in a variety of conversations, asking for details about their destination, suggesting a new dinner recipe, or answering a complex question without having to look at any screen or use any device.
"We aim to support conversations with natural dialogues and follow-up questions. Our customers can rely on us to ensure the best possible protection of their data privacy, said Markus Schäfer, member of the Board of Management of Mercedes-Benz Group AG, and chief technology officer - development and procurement.
Euphoria Turning Into Real-World Cases
With the rapid transformative potential of generative AI, the automotive industry is poised for significant advancements and disruptions alongside numerous other industries. A Meticulous Market Research report says that the automotive AI market is predicted to grow at a CAGR of 39.8% from 2019 to reach $15.9 billion by 2027. Another analysis from MarketResearch.biz says generative AI in the automotive sector is expected to be worth around $2105 million by 2032 from $271 million in 2022.
Similar to Mercedes, many other original equipment manufacturers or OEMs in the automotive industry are planning to leverage the potential of generative AI. German automaker BMW recently announced its adoption of generative AI to create customized car experiences. This innovative approach entails suggesting music and climate controls tailored to individual driver preferences, as well as providing customized navigation instructions.
General Motors is said to be working on a virtual personal assistant. This voice chatbot will also be based on machine learning models that power ChatGPT and will use Microsoft's Azure cloud service. While the company hasn't yet made a public announcement, Scott Miller, the company's VP of software-defined vehicle and operating system, said, "GM is developing an AI assistant, which could push things beyond simple voice commands."
Even Toyota has decided to discontinue support for Amazon's Alexa in its selected models and is working on improving its in-house voice assistant, including by potentially integrating OpenAI’s ChatGPT.
Generative AI is also finding practical applications in the realm of autonomous vehicles. A few months ago, China-based autonomous driving company Haomo.ai launched its large model DriveGPT. Haomo.ai will team up with partners to apply this to intelligent driving, driving scenarios identification, driving action verification and even provide assistance to the driver to move away from any complicated situation. Because of the volume of data, generative AI provides the capability of creating more practical algorithms to better train decision-making models.
Another interesting use case of generative AI comes from a Los Angeles-based EV manufacturer Faraday Future. While the delivery date of the carmaker is pushed, it has announced adding generative AI to its first car FF 91 Futurist. Its generative AI product stack, which will be added to the FF 91, will integrate its foundational AI capabilities with advanced models such as ChatGPT, GPT-4 and other similar large language models.
Generative AI Outside the Cockpit
As virtual and augmented reality becomes more mainstream, the application of generative AI sees another dimension. It will help create virtual environments in real time, providing an immersive experience. European auto OEM Fiat and Korean auto manufacturer Kia Motors are in the early stages of experiments to revolutionize shopping experiences through ChatGPT-integrated digital showrooms in the metaverse. The Fiat metaverse store has collaborated with Microsoft and software vendor Touchcast where customers can interact with an AI avatar "Fiat Genius" that answers questions about cars in real time with the help of ChatGPT.
Kia Motors Germany has also launched metaverse dealership based on a third-party metaverse platform Engage. Using their digital avatars, customers can walk around Kia's virtual showroom, interact with the vehicles and get their questions answered about the car.
Auto OEMs can also better the user experiences using generative AI. It can bring realism to virtual worlds for AR displays, user interfaces and language-aware voice assistants.
Big Tech Catching Up Fast With Demands
Technology companies, especially cloud providers and chip makers, had long sensed the opportunity and are already out with offerings to the auto OEMs.
Google Bard, the conversational AI language program that uses Google's existing LaMDA software, helps in offering advice to consumers when they are researching cars. It can offer very specific information like fuel type and even minor specifications.
Nvidia Omniverse - a dominant tool in the automotive sector - can build customized factories and shop floors. Omniverse even has the capability of creating a digital twin of the car that allows the customers to not only customize the car but also get their queries addressed during the process.
The automotive industry has begun adopting software-defined vehicles (SDVs) in order to offer a flexible and responsive solution for customers. SDVs have the ability to update and upgrade vehicle features through over-the-air updates. Generative AI can be used to create and optimize software and control systems, which improves the performance of the vehicle's hardware. Amazon's CodeWhisperer is an AI coding companion that uses generative AI to help improve developer productivity by generating code suggestions in real time.
Microsoft Azure has partnered with XPENG, a Chinese EV manufacturer to enable AI voice experiences for the automotive sector. This gives a fresh take on text-to-speech, expressive voice and self-service customization.
Conclusion
Despite having numerous use cases, the auto industry has barely scratched the surface of generative AI's potential. There are numerous other areas that are being experimented with. For example, generative AI can be applied to better the predictive maintenance and detect anomalies that could result in reduced downtime and direct cost benefits. It can also be used for optimizing supply chains.
It is true that generative AI requires enormously high computational resources and data, which can be expensive and resource-consuming. To run generative AI fundamental models, the users will need robust and responsive infrastructure that is purpose-built for machine learning. The issues of AI ethics cannot be neglected. However, the technology promises to be a transformative force for the auto industry, disrupting the way products are designed, manufactured and experienced.