AI is ubiquitous – but how do you make the best use of it? A short guide for suppliers

Munich, July 2024

AI is ubiquitous – but how do you make the best use of it? A short guide for suppliers

Munich, July 2024
S

uppliers need to exploit the opportunities offered by AI in the short term while realistically assessing the added value in the long term

The enthusiasm for artificial intelligence (AI) in the automotive industry is unbroken in the media, and management expectations regarding the potential to cut costs and therefore boost earnings through AI are high – and rightly so. This is also the case for the world’s 100 best automotive suppliers. Most of them take a public stance on AI – artificial intelligence is a topic in over 5% of all media coverage of the most active companies (see chart 1). Harman, a leading company, even mentions the topic of AI in more than 10% of all its relevant reports.

However, the dynamics of recent years also show that in economically challenging times, investments in the “future topic of AI” always involve risks – on the one hand, the risk of underestimating the significance of AI, but more serious, however, is the risk of being overly cautious in this area. This article is a compact guide to help you successfully navigate the use of AI in the supply industry by answering a few simple questions: What needs to be done and what should not done? Where is it worthwhile setting priorities to ensure a successful long-term competitive position, and where is technological innovation “simply a must”?

Key question: what needs to be done?

Our experience shows that the automotive companies leading in the use of AI have acted swiftly and have been determined to optimize business processes through AI and actively seize the available opportunities. Their efforts have focused on boosting efficiency “throughout the process” and “within the product.”

Throughout the process” includes optimization at all points that are suitable for machine optimization due to quantity, duration, effort, and/or complexity. On the one hand, these are “expected” tasks, such as the typical manually time-consuming creation and checking of specifications by AI – completion, pattern and error recognition, or translation can both improve quality and increase speed. On the other hand, AI can also be suitable for “unexpected” use cases, such as using it to help reduce electricity costs in production by detecting leaks in compressed air lines. The spectrum is therefore very broad and prioritizing the right cost-benefit ratio in line with the individual company’s position is crucial to success.

Although artificial intelligence “within the product” is usually associated with personalization within the vehicle, it also has its raison d’être in individualized customer interaction between or with supplier companies. Discussions with sector participants show that the focus here is on providing customers with initial information, which is achieved, for example, through context-sensitive digital agents and chatbots. From a supplier’s point of view, this may seem to play only a minor role, but this is precisely where undiscovered opportunities can lie. Low-threshold access to product and order information via an intelligent chatbot agent can help new customers in particular in terms of sales and at the same time scale up very efficiently without the need for additional staff. What’s more, ideally, AI-supported sales tools can be based on existing product data with little effort and create transparency regarding user behavior by analyzing the inquiries made. This in turn provides a more sound basis for making decisions on incremental improvements. Therefore, AI investment “within the product” can also pay off in the medium term for supposedly non-digital product portfolios.

Anteil der Artikel, in den KI erwähnt wird – Data frame: from 2023 (last 18 months)

Data Source basis: more than 100 sources, including InvestorPlace, Zacks Investment Research, Seeking Alpha, Reuters, CNBC, Market Watch, Forbes, Bloomberg Technology, Yahoo etc.

Follow-up question: what should not be done?

The use of AI always goes hand in hand with a discussion about the risks involved. Reservations about the secure handling of confidential and competition-related data, commercial concerns about initial development costs and long-term “lock-in” effects, but also ethics and bias justifiably play a major role. This is because the nature of (generative) AI sometimes makes it more difficult to uncover errors in results or detect the unconscious biases and unwanted behavior of an AI model. However, automotive suppliers should not allow themselves to be paralyzed by this factor.

It is essential to build the use of technology on a solid foundation. A key cornerstone of this foundation is choosing the right mix of AI partners who – in addition to in-house experts – can play an important and efficient role in implementing AI solutions. A competent technology partner can help to directly address the fundamental risks mentioned above. Particularly in the field of digital agents, the technology partner’s business model is often based on the customer’s ability to make successful use of AI. This means that high-quality results and data security, as well as the associated long-term use of a digital agent, are in the direct interest of the technology partner.

A look at the market shows that this strategy is bearing fruit, as due to the high speed of technological development, suppliers are increasingly willing to try out “fast-moving start-ups” instead of simply using established technology service providers. Particularly in application scenarios that require a quick ROI, start-up solutions of this kind can present a worthwhile alternative for automotive suppliers.

Final question: where should automotive suppliers use AI to make a conscious effort to improve their competitive position in the long term?

Despite the current challenges facing the entire automotive industry, the constant battle for the best competitive position will not be decided in a sprint, but in an endurance race, even with AI. The right talent is an essential ingredient. Automotive suppliers should therefore not only review the relevance of AI in their own company, but also aim to expand their own capabilities when it comes to analyzing, developing, and scaling AI solutions and thus create an attractive environment for AI pioneers and enthusiasts alike. This is an essential prerequisite for ensuring that AI solutions are perceived as more attractive by the company’s own workforce and therefore deployed to a greater extent. Only when this is successful can the hoped-for potential for cutting costs and thus boosting earnings come to fruition.

Here too, a look at the market allows us to draw an optimistic conclusion. According to our analysis, automotive suppliers that have increasingly leveraged (generative) AI during the last 18 months have consequently also experienced significant growth in their AI expertise. The last 100 employees to join the largest and/or most AI-focused automotive suppliers have industry knowledge from relevant companies. Therefore, from the perspective of these suppliers in particular, nothing should stand in the way of this trend.

Transfer of technical staff – Sankey diagram with pure tech companies, excluding tech consulting companies

Source: Berylls by AlixPartners

Autoren
Malte Broxtermann

Partner

Clinton Charles

Senior Data Scientist

Malte Broxtermann

Malte is an expert in the development and implementation of automotive digitization strategies.

He focuses on helping clients scale (generative) artificial intelligence to improve their bottom line across the entire automotive value chain. His primary customers are automotive manufacturers and their suppliers, especially those active in the Software-Defined-Vehicle space.

Before his time at Berylls by AlixPartners (formerly Berylls Strategy Advisors), he advised leading North American utility companies. Prior to that, he saved lives as emergency medical technician. Malte holds master’s degrees in economics from Maastricht University and Queen’s University in Canada.