Contributors: John W. Schmotzer, Head of Automotive Ecosystem Business Development, NVIDIA and Guy Bursell, Head of Business Strategy, Manufacturing and Supply Chain, Microsoft
The automotive industry continues to adopt artificial intelligence (AI) at an increasingly rapid pace to keep abreast of numerous production challenges and market opportunities. AI offers many advantages in engineering and manufacturing, from speeding up design iterations to improving quality control on production lines. AI is also helping forward-thinking automotive companies adapt efficiently and quickly to rapidly changing business and industrial priorities.
Map data to the right destination
Manufacturing is the foundation of a strong automotive business, and high-quality data is the foundation of a strong AI strategy. The data landscape used to be static and historical, but now it is also streaming and informative in real time. Managing data from a variety of formats and sources is a challenge, and it’s increasingly difficult to keep files and datasets properly labelled, merged, cleansed, and up-to-date.
But it’s not just the pressing needs of managing large and changing data sets that are testing the limits of manufacturing IT teams. Changes in data usage and storage affect IT infrastructure and AI performance. The continuous flow of AI models, decision rules, and HPC projects affects data demands and the processes that data powers. In modern manufacturing, data presents more challenges beyond its ever-increasing volume, variety of formats, and seemingly endless points of origin.
“The highway to AI success relies on integrating disparate data sets that have not traditionally been combined to form new insights identified through the use of sophisticated AI technologies,” said John W. Schmotzer, Director of Automotive Ecosystem Business Development at NVIDIA.
As data management becomes a top priority for the automotive manufacturing community, several opportunities and challenges present themselves. One of the opportunities for streamlining manufacturing processes is the ability to merge data sets from various organizations across the business to provide a single view of the vehicle throughout its lifecycle. This can reduce warranty accrual costs and streamline material costs year over year. Warranty accumulation alone averages $600 per vehicle in additional costs for major automotive OEMs and can be a deciding factor in influencing customer loyalty.
New opportunities for cost optimization create new challenges for manufacturing companies. It could be difficult to consolidate manufacturing data, connected vehicle data, metadata, and CAE rendering files from across the enterprise into a single view and singular format. Schmotzer calls it the “Recursive data lake problem” due to nested information kept in different organizations and sub-organizations within the company.
Strategic management decision-making on software tools and processes that standardize agreed formats and ways of working will help alleviate these bottlenecks while pursuing Industry 4.0 opportunities.
Building the Highway to AI Success
the advent of connected battery electric vehicles and AI for visual inspection, we are on the cusp of a truly transformative manufacturing experience. The term ‘Industry 4.0’ does almost a disservice to the scale of change and the real business value being realized in transportation,” says Schmotzer.
While AI proves to be a crucial tool in meeting almost any production demand, such projects rarely operate at maximum capacity. AI puts great pressure on resources, but upgrading a manufacturing plant to a smart factory can reduce impact, stretch resources, optimize processes, improve efficiency, ensure product quality, speed up R&D results and reduce maintenance costs.
For example, scaling up accelerators and interconnected accelerator networks can dramatically improve inference, AI training, and model-parallel training needs, among other uses of model building and modeling. AI training. Built-in tool chains for a variety of user skill levels can take advantage of different user abilities and also democratize AI projects. Support for scaling Machine Learning Operations (MLOps) solutions is also essential for increased productivity. Built-in features and metrics to ensure Responsible AI can go the extra mile in preventing future issues, including potential product recalls and liabilities, among other issues.
Learn how Microsoft Azure and NVIDIA give BMW the computing power for automated quality control.
Deep learning (DL) has many advantages over its less resource-intensive cousin, machine learning (ML). Preserving the resources consumed by deep learning projects is key to making the use of technology sustainable. GPUs are unique and highly capable of solving DL problems due to their high power efficiency and excellent price performance. GPU-powered computing is an excellent choice for highly parallelized environments and large-scale repeatable tasks.
“Cloud infrastructure helps manufacturers break free from the chains of limitations and constraints. HPC and AI in the cloud empower manufacturers to ask larger, more complex questions, the answers to which increase impact, innovation and differentiation in crowded markets,” said Guy Bursell, Chief Strategy Officer sales, manufacturing and supply chain at Microsoft.
Trace execution path
Pinning AI to the competitive realities of automotive manufacturing is crucial. Businesses can meet the demand and scale of customer needs by modernizing into a smart factory, enabling optimized processes, increased efficiency, assured product quality and reduced maintenance costs.
AI-driven toolchains and cloud infrastructure can have a significant impact on a manufacturer’s bottom line. Microsoft and NVIDIA’s comprehensive cloud infrastructure, purpose-built for AI, delivers real-time speed, predictability, resiliency, and durability that can help businesses accelerate innovation and IT workloads. AI.
Learn more about Microsoft Azure and NVIDIA solutions for AI:
#Building #Highway #Success #Cloud #Infrastructure