Viaduct, a startup based in Menlo Park, Calif., is aiding vehicle and parts manufacturers by developing models for time-series inference to extract failure insights from connected car sensor data.
The company employs NVIDIA A100 Tensor Core GPUs and the NVIDIA Time Series Prediction Platform framework to detect anomalies, track issues and deploy failure predictions, contributing to significant cost savings and improved quality control across the automotive industry, according to an Aug. 15 blog post.
"Viaduct has deployed on more than two million vehicles, helped avoid 500,000 hours of downtime and saved hundreds of millions of dollars in warranty costs across the industry," David Hallac, founder and CEO of Viaduct, said in the post. "Customers see it as a huge savings — the things that we are affecting are big in terms of profitability."
Viaduct, founded four years ago, specializes in crafting models for time-series inference, thereby aiding enterprises in extracting failure insights from the data harnessed by contemporary connected cars. This is achieved through a strategic utilization of sensor data, fostering correlations to unearth meaningful patterns, the blog post reported.
At the heart of Viaduct's endeavor lies its offering as a dynamic platform facilitating the detection of anomalies, effective issue tracking and the deployment of predictive failure scenarios. This innovation empowers automakers and parts suppliers to proactively address challenges by leveraging real-time data, the post said. The outcome is a notable reduction in warranty claims, recalls and defects.
The technological bedrock underpinning Viaduct's prowess lies in the sophisticated fusion of NVIDIA A100 Tensor Core GPUs and the NVIDIA Time Series Prediction Platform framework, the post noted. These elements synergize seamlessly to underpin the training, fine-tuning and deployment of time-series models that furnish insightful forecasts based on data trends.
Viaduct's eminence resonates across the industry, as it has partnered with more than five prominent manufacturers spanning passenger cars and commercial trucks.
Viaduct, a proud member of NVIDIA Inception. The program offers technological support and AI platform guidance to enterprises, and traces its origin back to Stanford University, according to the blog post.
Hallac, during his Ph.D. tenure, embarked on a journey catalyzed by sensor data provided by Volkswagen, alongside a research grant. The synergy aimed to unravel the potential applications encapsulated within voluminous vehicle data over extended periods, the post reported.
