EV adoption has been slow as consumers continue to have doubts, so automakers are stepping up their marketing efforts to counter the myths.
IAA Mobility 2023, the world’s largest mobility event, drew around half a million visitors to Munich from September 5 to 10. Artificial intelligence (AI) in the auto industry was a recurring theme in almost all areas of this year’s show. For example, in the development of driver assistance systems and autonomous driving, or in quality control and production.
In automotive factories, AI-controlled robots are now independently performing tasks such as welding, painting and assembly.
Increasingly, intelligent algorithms are also being used to monitor the condition of vehicles and provide indications of upcoming maintenance or repairs, also known as “predictive maintenance.”
Artificial intelligence is also used in vehicle design and to optimize driving for greater efficiency and sustainability in voice control of navigation systems and in intelligent parking aids. Meanwhile, marketing, sales and customer service implement AI to make customers happier and supply chains more efficient.
🤩 Shine Bright: 🤩
Real-life transformation – thanks to AI-powered chatbots + digital self-service technology.
Get the details HERE.
AI in the auto industry: How artificial intelligence is changing the course
I had the pleasure of discussing AI at the Mobility Festival, and was joined by Alexander Scholz, Head of Digital Supply at BMW Group, as well as Tobias Wagner from the e-mobility startup ChargeX. Amidst the hustle and bustle of the trade show, we were able to use the Executive Lounge, operated by our partner IBM iX DACH together with TikTok, for an exciting AI MasterClass.
Generative AI is a real game changer, especially in the area of communication.
That’s because this technology can generate new content based on existing information and user input. It’s based on large language models (LLMs) and is used in AI tools such as ChatGPT, Google Bard, and Aleph Alpha. When trained on large amounts of data in many different contexts and dimensions, such machine learning (ML) models can now understand complex relationships and dependencies.
For BMW expert Alexander Scholz, this technology is also an important efficiency driver, especially in the supply chain.
The benefits of AI in the auto industry are already being felt in production. At BMW’s US plant in Spartanburg, for example, the use of AI in the body shop alone is saving more than a million dollars a year in production costs. And the company is already experimenting with artificial intelligence in vehicle design, for example by designing new off-road vehicles without human intervention.
The young company ChargeX also relies on an AI solution for its modular e-car charging infrastructure. It’s used to automatically distribute the load between the various electric cars at a site. “We can use it to develop an optimal charging strategy,” founder and CEO Tobias Wagner told us. But they are still in the early stages.
Keeping an eye on the potential risks of AI in the auto industry
Despite the different perspectives, the discussion also revealed many commonalities. For example, when we talked about the potential risks of AI – such as data security, protection of sensitive information, or liability and warranty issues.
“We need to be proactive and ensure the greatest possible transparency,” Scholz emphasized. He added that it’s important to use AI language models responsibly and to build trust in their use among one’s own employees and customers.
For this reason, BMW has already published their own AI guidelines, which set out the ethical principles for dealing with this disruptive technology. These include not blindly relying on AI responses without human control.
In order to prevent the AI from “hallucinating” – especially in safety-critical situations – it must be ensured through appropriate training that the output of an LLM is factually correct and unbiased. Furthermore, in case of doubt, the judgment of a human should always prevail over that of an AI.
Without end-user acceptance, the best AI solution is useless
Tobias Wagner brought another aspect into our discussion: the absolute necessity of end-user acceptance.
He said that the automotive industry needs to be particularly sensitive to this, because drivers want to make their own decisions, not leave them to an opaque algorithm.
He pointed to his company’s charging app, which in an earlier version automatically determined the optimal charging process for the electric car based on historical data and the current situation at a specific location.
“But people want to decide for themselves, based on their specific situation, how full their battery should be and how much time they have to do it,” he said, speaking from ChargeX’s experience. Reasonable suggestions and recommendations from the AI are helpful, he said, but the ultimate decision must rest with the customer.
Technology has changed how we shop for everything, including car buying. Find out how two companies are transforming the automotive industry.
Rather than regulate AI, it’s better to make your own experience
All panelists agreed, however, that permanent regulation of the new technology – of any kind – would be of little help. It would only slow down innovation, and Germany would fall behind in another area.
However, it’s often necessary to discuss in detail with the legal departments of car manufacturers what’s currently possible for reasons of liability or data protection, and where the limits might still lie. Working through the issues together should ensure that the implementation of new use cases is not delayed due to legal concerns or existing fears.
Gathering employees’ own experiences can also help allay serious concerns about AI applications.
The experts in the MasterClass were convinced that “AI development can no longer be stopped or even reversed.” Rather, the challenge is to shape it and use it responsibly.
Without data quality, AI will not spit out useful results
At our MasterClass in Munich, we also discussed another key issue in the use of generative AI in the automotive industry: The collection and structuring of data and its quality. It was noted that the best analysis tools are useless without quality data – if the data is poor, then even the best AI can only provide useless answers.
If, on the other hand, the relevant information is collected and analyzed at all touchpoints along the customer journey, customers can be sent customized offers via their preferred channel, for example. However, this requires their consent.
Our panel agreed that generative AI is the ideal tool for personalized marketing that is tailored to the recipient’s current situation. This is because it’s an excellent way to automate 1:1 campaigns that are highly relevant to customers and therefore produce better results.
Two-step approach to developing AI solutions
According to Scholz, BMW is pursuing a two-stage approach to the further use of artificial intelligence. The first step is to use it to increase efficiency in all areas, reduce workload and relieve employees of routine tasks. In the second stage, it will be easier to make more precise and better decisions based on the data collected. This would also provide effective support for employees in the face of demographic change and increasing staff shortages.
Our discussion on AI in the auto industry can be summarized in the following points:
- There are already an infinite number of use cases in the automotive industry where AI can be put to good use. We discussed some examples in our session, ranging from supply chain to charging infrastructure to customer-facing processes – but we’re only at the beginning of the evolution.
- The technology is changing very fast. Therefore it’s worthwhile for the industry to set up dedicated teams / competence centers in their organization to keep an eye on developments and to be able to react quickly to new trends.
- The current trend is for automotive companies to have their own “enterprise ChatGPT” customized to their specific needs and trained with their own data to guarantee the quality of the results.
- Clean data is the key to getting interesting results from AI deployments from a business perspective and improving the customer experience.
- One of the biggest challenges today is finding employees with the necessary AI skills or training them yourself.
- Good and transparent communication is essential to address and hopefully allay the concerns of employees and customers.
It was enriching for me to hear first-hand from our experts on the panel how they are using AI to make their own business processes more efficient. But also how they’re using it to improve sales, marketing, and service, and most importantly, to create a better customer experience for their customers.
It’s an exciting time, and I’m very excited to see what’s next for AI in the automotive industry. Are you?