From autonomous driving to manufacturing Generative AI: Game-Changer in the Automotive Industry

Related Vendor

Wuxi InfiMotion Technology Co., Ltd.

Whether it's autonomy or electromobility in passenger cars, commercial vehicles, or off-road sectors: With generative AI, generative design becomes possible, enabling significant weight savings of 15 to 50 percent depending on the application, thus allowing for more sustainable products that also meet cost requirements. Parameters are optimized with AI over countless variations, and innovation cycles are significantly shortened. AI becomes a game-changer in the automotive industry.

*Ralf Klädtke is Chief Technology Officer (CTO) at TE Connectivity for the Transportation Solutions segment.

The entire automotive industry is under high pressure to introduce new propulsion technologies and software-defined vehicle platforms, to implement sustainability across the entire lifecycle, while at the same time reducing costs, increasing efficiency, and above all, significantly shortening the time-to-market. The so-called "squaring of the circle" is the new "New Normal" for the automotive industry, which has already led to several insolvencies. With the use of generative AI, a paradigm shift is currently taking place that will fundamentally change the automotive industry.

What potential does AI hold?

According to Goldman Sachs Research, recent successes in the field of artificial intelligence (AI) have the potential to bring about far-reaching changes in the global economy. "With the use of tools that harness natural language, the global GDP could rise by about seven percent, or seven trillion US dollars, over the next ten years," their report states. McKinsey is similarly optimistic. Their analysts believe that due to AI-induced productivity growth, the global economy could grow by up to 4.4 trillion US dollars annually. In their examination, McKinsey looked at 63 use cases and found that three-quarters of the growth is attributable to the four areas of marketing & sales, customer service, software development, and research & development.

Especially affected: The transport and automotive industry

AI in the transportation and automotive industry is nothing new. Long before the hype around ChatGPT, there were already many AI solutions, mainly in the form of predictive models, for example, in predictive maintenance, in image processing for quality control, or for transport optimization in logistics. Above all, AI is used in the area of autonomous driving to reliably detect objects with different sensors such as cameras, LiDAR, or radar under all weather conditions, thus making autonomous driving safe.

The introduction of generative AI (GenAI) is causing a quantum leap in nearly all sub-sectors of this industry. This technology optimizes practically all aspects of the automotive industry: vehicle performance, vehicle development, sustainability, transportation and logistics, marketing and sales; and it also gives new momentum to autonomous driving. GenAI creates millions of realistic scenarios and, in particular, hazardous situations with which the AI for autonomous driving is trained. GenAI trains the AI of autonomous driving. Currently, it is expected that over 90 percent of the many years of training for autonomous driving and millions of training kilometers can be virtually trained through generative AI. According to market researchers from Precedence Research, the global market for GenAI in the automotive industry reached a value of 312 million US dollars in 2022. Last year, it is estimated to have been about 2,700 million US dollars – almost nine times as much. For the period up to 2032, analysts expect an average annual increase (CAGR) of 24 percent.

Interplay of human and AI

Besides autonomous driving, another important focus of GenAI utilization is in research and development. Generative AI is trained with companies' own confidential data, enabling the development of new design concepts in the shortest possible time, conducting the relevant simulations, and also creating the necessary phase documents. Brilliant innovations are not to be expected from AI in this regard. However, variations of existing product families are within the realm of possibility.

Generative design, on the other hand, is a 3D CAD feature that uses AI to translate system requirements into autonomously optimized designs. The AI iteratively calculates countless variants and optimizes the design to still meet all mechanical, thermal, or other requirements. The resulting generative design is a collaboration between human and AI, leading to extraordinary and sustainable design solutions that a human alone could not develop. Design cycles become significantly faster and more efficient.

According to Precedence Research: "The use of GenAI in the auto parts sector can reduce the cost and time expenditure in development by up to 20 percent," their report states. This means that, with the help of GenAI-based optimization of design, manufacturers can generate unique selling points while significantly reducing production costs and time-to-market. With generative design capable of quickly generating and evaluating countless iterations, the traditional constraints of time and resources are overcome. This allows the automotive industry to explore new solutions that were previously unthinkable. Companies that employ AI will have significant competitive advantages compared to companies that develop traditionally.

GenAI plus 3D-Printing equals reduced development times

TE Connectivity has developed its own strategy for GenAI. It has examined how AI will influence the industry, business model, products, and also processes. They have defined AI ambitions and use cases that will be introduced in the future. This will take time but will encompass the entire value chain and all functions within the company. In pilot projects, TE Connectivity has improved the Product Carbon Footprint of products by over 50 percent with generative AI design, thus achieving significant benefits for sustainability. The use of AI saved over 90 percent of the time for simulations of tools in injection molding while simultaneously optimizing cycle times. AI was used to optimize multiple parameters simultaneously, thereby improving product design.

Work is being done on the vision that with GenAI and 3D printing, the product development times for certain products can be reduced from the current two years to just 14 days. If customers have system requirements that do not require significant new developments, AI is already capable today of creating variations of existing products in an extremely short time based on existing, company-confidential data. By building Digital Twins, concept variants can be developed quickly and effectively based on company-confidential existing test data and AI simulation data. GenAI in the form of Large Language Models (LLM) is also capable of creating phase documents based on the generated data, thereby accelerating the entire development process and creating efficiency that was unthinkable until recently. Artificial Intelligence is capable of creating a concept design in the shortest possible time, and just a few days after the initial customer inquiry, a first prototype can be delivered using 3D printing. Certainly, in real implementation, the AI results need to be verified and validated through tests and simulations, since AI is trained purely with data and indeed also has numerous limits. Humans also play a central role in the use of AI. AI is a tool for humans, which must be used intelligently to become faster and more productive. The rapid development of AI in combination with trained employees who understand how to use AI intelligently will open up entirely new possibilities for transformation across the entire value stream of the industry.

GenAI: From strategy to beneficial use

Before a company can flawlessly play on the keyboard of GenAI and the adjacent technologies, important stages must be completed. It begins with developing an AI strategy, where AI capabilities are incrementally increased. This starts with simple experimentation with GenAI to develop a basic understanding of the possibilities and also the limits of AI. After that, it is vital to identify use cases that are feasible and generate real added value for the respective company. Cost-benefit analyses are helpful in prioritizing. In parallel, it is then necessary to empower the organization, train employees, and create expertise and culture within the company to successfully deploy AI. The maximum value enhancement through artificial intelligence is then achieved when the entire company's DNA has been transformed across the value stream.

Human factor necessary

One thing is clear: The transport and automotive industry is facing a profound change in the working world, where development and production processes will fundamentally change due to the use of GenAI. This means that new tasks will emerge that existing employees will need to learn and carry out. It also means that new AI positions will be created that did not exist before. AI, data, and machine learning developers will be just as needed as data analysts to professionally handle the new tasks. Only in this way can the numerous new AI possibilities be implemented in the interest of company success. It is especially important that GenAI is perceived not as a threat but as an opportunity. GenAI, combined with trained employees, makes a company successful in competition and maintains and creates jobs. An important insight is also that GenAI is "not really intelligent". A GenAI trained with incorrect data will also deliver incorrect results. The combination of intelligent employees and GenAI is the model for success.

Computing platform Drive

Nvidia: Extensive collaboration on autonomous driving

Conclusions

In an industry characterized by the New Normal of disruptive technological innovations, high sustainability demands, ever-shorter development cycles, and simultaneously high cost pressures, the use of GenAI heralds a new era. GenAI in the transport and automotive industry is a paradigm shift that will change the entire value stream of companies. Electromobility and autonomous driving are just small steps into the future of mobility, which will be shaped by technological innovations, sustainability, and safety. Fleets of autonomous, electrified agricultural machines such as combine harvesters, excavators, and also trucks will perform many tasks autonomously and realize productivity for people. In megacities, Mobility-as-a-Service will be increasingly utilized. Robotaxis and autonomous shuttles will be available to people in inner cities. Private vehicles will no longer be parked for 95 percent of their lifespan in megacities, and autonomous safety will lead to a drastic improvement in road safety. Human creativity and intelligence combined with machine learning, neural networks, and GenAI will make this future possible.  (se)