Automakers Extend AI Use Beyond Autonomous Vehicles into Back Office

Automakers, suppliers and startups developing self-driving cars are implementing deep learning algorithms to operate vehicles without a human driver. With high-performance compute solutions from NVIDIA, vehicle manufacturers are also using AI in their data centers to enhance day-to-day operations.

From predicting inventory demand to training algorithms to make an unprotected left turn, automakers and NVIDIA DRIVE ecosystem partners are turning to high-performance compute solutions to solve the problems of today and tomorrow.

Read More: Interview with Ben Goertzel, CEO at SingularityNET

Meeting Customer Needs with an AI for Detail

When it comes to selling cars, time is money. Extra inventory sitting on lots can cost automakers and dealers thousands of dollars, while cars with initial quality issues can lead to costly repairs and disgruntled customers.

With the ability to learn from past patterns, AI can help ease both these bottlenecks in the sales process. Using NVIDIA GPU-accelerated computing solutions to run predictive deep learning algorithms, Volkswagen can project demand for specific models.

The result is fewer vehicles sitting idle at dealers and higher customer satisfaction. Volkswagen is implementing a similar AI process in its on-demand concierge service to predict and efficiently meet customer needs at time of service.

Read More: Palo Alto Networks Appoints Amit K. Singh as President

The automaker is also leveraging NVIDIA AI infrastructure to ease customer pain points throughout the vehicle life cycle. By ingesting traffic data to build an AI model that predicts the speed of traffic and road occupancy,  automakers can provide drivers with congestion updates and adjust the engine to operate most efficiently for upcoming road conditions.

New vehicles that aren’t up to quality standards can take a significant chunk out of an automaker’s bottom line. In the past five years, carmakers have set aside an average 2.7 percent of sales revenue every time they sold a vehicle, totalling $50 billion in w******* funds annually.

At a panel on datacenter strategies at GTC Europe, BMW said it’s addressing this challenge by using NVIDIA-powered deep learning algorithms to identify and address quality issues before cars get shipped — a more efficient process than visual inspection and traditional monitoring activities.

Read More: Interview with Angel Gambino, CEO and Founder of Sensai

Efficient Training for AI Drivers

While data center solutions for AI are easing auto industry inefficiencies, companies are also using them to prepare for a new automotive future.

Autonomous Intelligent Driving (AID), a subsidiary of Audi developing self-driving technology for VW, Audi and Porsche, as well as BMW, are relying on NVIDIA GPU-accelerated computing solutions to train deep learning algorithms for autonomous driving.

Read More: Interview with Jeffrey Kofman, CEO and Founder at Trint

Zenuity, a joint venture between Volvo Cars and supplier Veoneer, is leveraging NVIDIA DGX for self-driving AI training. With the high-performance datacenter, the autonomous driving software supplier is able to handle the exponentially increasing volume of sensory data to train their models, reduce data bottlenecks and speed up development cycles.

The high-performance data center solution enables companies to use massive amounts of data to train AI at unprecedented speeds. Engineers can use camera and sensor data collected from vehicles to teach AI algorithms the rules of the road and how to react to various traffic scenarios.

Read More: The Top 5 “Recipes” That Give AI Projects a Higher Likelihood of Success

This streamlined solution brings safe self-driving to roads sooner, mitigating the $500 billion lost in damages due to traffic accidents each year.

These uses are just a few examples of how AI infrastructures can transform the industry, from new driving technologies to more efficient operations. With the level of compute enabled by NVIDIA data center solutions, the possibilities to work smarter and faster are endless.

Read More:  Fluor Uses IBM Watson to Deliver Predictive Analytics Capability for Megaprojects