There is a significant amount of excitement surrounding the transformative potential of artificial intelligence (AI), with a consensus among many experts that the critical factor in harnessing this emerging technology lies in identifying the appropriate business use case.
In a one-on-one video interview with ZDNET, Thierry Martin, the senior manager for data and analytics strategy at Toyota Motors Europe, delves into how the automotive behemoth is investing time and resources in research and development to leverage the capabilities of AI.
The current focus of their exploratory efforts primarily revolves around data science rather than prediction and automation. Martin emphasizes the importance of data analysis for Toyota, citing examples such as examining how people drive their cars and discerning any disparities between driving behaviors in different regions like Germany and Belgium.
Toyota’s proficiency in data collection enables them to gain valuable insights into the usage patterns of their vehicles. This insight is instrumental in various aspects, including forecasting models for root cause analysis and predicting necessary accessories for planning purposes.
While Toyota currently prioritizes maintaining human involvement in decision-making processes by using tools like Power BI to enhance analytics and operational understanding, Martin envisions a future where AI will play a more prominent role in production. Efforts to identify suitable use cases for business processes are already underway within the organization.
Looking ahead, Toyota is exploring avenues to incorporate generative AI securely within their operations, with a particular interest in analyzing text data and leveraging models like ChatGPT. Despite the advancements made by OpenAI’s large language models (LLMs) in popularizing generative AI, Toyota exercises caution in deploying these cutting-edge technologies, aligning with the approach adopted by other leading enterprises like CyberArk.
Martin outlines two strategic pathways for maximizing the potential of AI at Toyota Europe. The first route involves leveraging tools such as Microsoft Copilot at an individual level to facilitate task completion using non-sensitive data. The second pathway entails exploring the secure utilization of generative AI within the enterprise firewall to drive productivity enhancements.
To support their AI initiatives, Toyota Europe has established a data mesh framework focused on governance and data product ownership. This framework is underpinned by technologies like Snowflake for data access governance, Dataiku for collaboration, Collibra for governance, and Denodo for connecting data meshes across different organizational units.
By harnessing these data mesh technologies, Martin’s team has developed chatbots on Dataiku, incorporating a language model that operates on a secure Azure Open AI instance to provide document summaries. The internal development of these AI tools helps address concerns associated with publicly available models from major providers, offering a safer and more controlled environment for experimentation and deployment.
As Toyota Europe continues to explore AI solutions, rigorous testing is paramount to ensure compliance with governance standards and the specific requirements of business users. Martin underscores the importance of ethical considerations in AI deployment, emphasizing the need for clear guidelines on data access and governance.
Looking ahead, Martin aims to expedite the deployment of AI-based tools into production while collaborating with technology partners like Snowflake, Dataiku, and Collibra to address governance challenges and data access restrictions. By mid-2024, Toyota Europe anticipates having AI-enabled chatbot services extracting data from the Snowflake platform, with ongoing discussions on realizing this vision with key technology partners.
Ultimately, Martin stresses the significance of close collaboration between technology experts, business leaders, and engineers to effectively demonstrate the value of AI services and identify optimal deployment scenarios within the organization. This collaborative approach is crucial in ensuring a comprehensive understanding of AI applications and maximizing their impact across various business functions.
