Guest commentary: Overcoming legacy data constraints to unlock the potential of generative AI

The automotive industry is on the cusp of a major transformation, driven by the development of autonomous vehicle systems and other advanced technologies such as software-defined vehicles. One of the key challenges facing automakers is the need to generate and validate massive amounts of data to train the artificial intelligence models that power these new technologies. Generative AI has the potential to revolutionize the way automakers develop and test their products. By using generative AI, automakers can create synthetic data that is indistinguishable from real-world data. This can help them train their AI models more quickly and efficiently, bringing new products to market faster.

At its core, generative AI represents a monumental shift in how data is harnessed. It offers a promise of accelerated data transformation, tackling issues ranging from code quality to resource efficiency. With its ability to automate routine tasks and streamline software development processes, generative AI presents an opportunity to enhance scale and agility while elevating the customer experience to new heights.

However, legacy data constraints threaten to impede progress. Decades of siloed data management, outdated IT structures and a lack of visibility across the supply chain have created barriers to innovation. Fortunately, these challenges aren't insurmountable; rather, they represent opportunities for growth and transformation.

The key is recognizing that generative AI isn't just a solution; it's a catalyst for change. By leveraging generative AI to address legacy data and tech debt issues head-on, stakeholders can unlock the technology's full potential. From code conversion to workload migration, generative AI offers a pathway to more effectively tackle long-standing challenges that have hindered progress across the automotive industry.

For instance, employees used to working with older mainframe technology are retiring from the workforce, and generative AI could be used to transition the data model into Python or move it to the cloud.

One of the critical areas where generative AI can make a significant impact is in improving code quality and resource efficiency. Traditionally, software development has been more art than science, with each developer employing their own methods and practices. Generative AI can change the game by offering repeatable, standardized solutions that streamline development processes and drive efficiency gains.

In addition, the ability to create realistic simulations using generative AI is a game changer. Imagine virtually testing a car's performance in extreme weather conditions or on treacherous mountain roads before a single piece of metal is bent. This can significantly accelerate development cycles and reduce the need for expensive physical prototypes.

Generative AI can also be a powerful tool for quality control. By generating synthetic data that includes specific defects, AI systems can be trained to become highly adept at identifying those same defects on the assembly line.

This translates to fewer missed defects, improved product quality, and ultimately, a safer driving experience for consumers.

There are valid concerns surrounding the use of AI-generated data to train critical vehicle systems. It's important to ensure that a human has oversight into the process and observes proper governance, compliance and security measures to protect against intellectual property infringements, while implementing rigorous validation processes to ensure the accuracy of the synthetic data.

Additionally, close collaboration with domain experts in the automotive field is essential to ensure the generated data reflects real-world scenarios. A recent trend known as "grounding" makes sure that generative AI data isn't being made up simply to provide a response. With grounding, the generative AI model ensures the answer provided is cited and links to actual company data.

The benefits of generative AI can extend beyond the manufacturing floor. It can significantly improve the customer experience by creating a unified journey of touchpoints across all interactions, from the auto manufacturer to the dealership.

Dealerships could also leverage generative AI to create personalized marketing campaigns or targeted service recommendations for individual customers based on their driving habits and preferences. Whether it's navigating a dealership website or seeking assistance with loan inquiries, generative AI promises a more cohesive and personalized experience for consumers.

Generative AI could also pave the way for personalized in-car experiences, tailoring everything from entertainment options to climate control based on preferences.

Realizing the full potential of generative AI requires an effort to address legacy data constraints. From modernizing IT infrastructure to fostering collaboration across the supply chain, stakeholders should confront these challenges head-on. Generative AI isn't a fad, and manufacturers should consider piloting programs using this technology now to avoid being left behind.

Gen AI represents a paradigm shift in automotive technology, offering both promise and peril. While its transformative potential is undeniable, it's imperative that stakeholders proactively address legacy data constraints to fully capitalize on its benefits. By embracing generative AI and tackling legacy challenges with urgency and determination, the automotive industry can chart a course toward a future defined by innovation, efficiency and customer-centricity.