Sure! Here’s the translation:
—
The collection of vehicle data has become an essential pillar for original equipment manufacturers (OEMs) in their efforts to innovate and enhance the performance of their products. With the increasing digitalization of vehicle architecture and the incorporation of software-configurable functions, OEMs can add new features more swiftly. In this context, the company Sonatus has launched two tools, Collector AI and Automator AI, designed to facilitate the transition to Software-Defined Vehicles (SDVs).
Collector AI promises to simplify data management throughout the vehicle’s lifecycle by implementing data collection policies that do not require alterations to the vehicle electronics or embedded software. Despite its advantages, engineers and vehicle data users face the challenge of choosing from thousands of available signals for each specific case. Meanwhile, Automator AI, which offers a no-code methodology for function automation, presents similar complications for those unfamiliar with the available signals and events.
To overcome these obstacles, Sonatus has established a partnership with the AWS Generative AI Innovation Center. Together, they have worked on developing a natural language interface to facilitate the creation of data collection and automation policies, all utilizing generative artificial intelligence. The goal is to reduce a process that traditionally could take days to mere minutes, making it easier for both engineers and individuals without a technical background to access.
To achieve this, Sonatus has implemented techniques that allow for the generation of policies from natural language inputs, significantly reducing creation time and expanding participation to a more diverse audience, including non-technical stakeholders. However, the process has not been without challenges; the complexity of vehicle event structures and the scarcity of labeled data have complicated implementation.
Among the established success indicators are the reduction in time needed for policy generation and the increase in the number of policies generated per client. A specialized advanced technology team, along with AWS, has developed an automated system that enhances efficiency by breaking down user requests into structured and actionable components.
This multifaceted approach emphasizes the precise identification of signals through vector-based methods and utilizes agents to adjust signal proposals, thus ensuring a clear and efficient outcome. The combination of adaptive work structures, parameter generation, and a contextual approach has revolutionized policy creation, resulting in not only increased efficiency but also a significant reduction in the time required for implementation.
With these innovations, Sonatus, supported by AWS, continues to lead the digital transformation and automation in the automotive sector, enabling organizations to manage complex workflows more effectively.
Referrer: MiMub in Spanish
