ARTIFICIAL INTELLIGENCE AUTOMOTIVE'S NEW VALUE-CREATING ENGINE
January 2018
15+
2 3 4500+
ARTIFICIAL INTELLIGENCE AUTOMOTIVE'S NEW VALUE-CREATING ENGINE
CONTENTS
Introduction and key insights 6 1. Mapping the new landscape of value opportunities for automotive OEMs ..........10 2. AI-enabled applications provide substantial new value opportunities for automotive OEMs ..........16 21 Industry-wide value opportunities: applying AI to processes 16 22 OEM level competitive advantages: customer-centric services20 3. Fully capturing the AI-enabled value opportunities requires OEMs to initiate an AI transformation........ 22 Getting started 24 Appendix: Methodology of McKinsey's artificial intelligence market model26 Legal notice 27 Contributors 28
Artificial intelligence: automotive's new value-creating engineINTRODUCTION AND KEY INSIGHTS
For more than two years now, the automotive industry has been intensively discussing four disruptive and mutually reinforcing major trends autonomous driving, connectivity, electrification, and shared mobility These ACES trends are expected to fuel growth within the market for mobility, change the rules of the mobility sector, and lead to a shift from traditional to disruptive technologies and innovative business models Artificial intelligence (AI) is a key technology for all four ACES trends Autonomous driving, for example, relies inherently on AI because it is the only technology that enables the reliable, realtime recognition of objects around the vehicle For the other three trends, AI creates numerous opportunities to reduce costs, improve operations, and generate new revenue streams For shared mobility services, AI can, for example, help to optimize pricing by predicting and matching supply and demand It can also be used to improve maintenance scheduling and fleet management These improvements through AI will play an important role for automotive firms because they enable them to finance and cope with the changes ahead of them One expected key result from the ACES trends is a marked shift in the industry's value pools This change will primarily affect large automotive original equipment manufacturers (OEMs) and their business models, but the impact will be felt throughout the industry and beyond The products and services made possible by the ACES trends will not only impact the business of all incumbent and traditional industry players, but will also open the market up to new entrants Many companies that were previously focused on other industries, eg, technology players, are heavily investing in the ACES trends and the underlying key technologies As a result, a new ecosystem of players is emerging New players will be important partners for traditional automotive companies While automotive OEMs can use new players' technology expertise to unlock value potential from AI, new players will have opportunities to claim their share of the automotive and mobility markets To master the ACES trends, OEMs need to invest substantially into each of the four ACES not just in their development, but also in their integration Against this backdrop, this article which is a continuation of our work1 on AI in the automotive sector and the insights of which are based on a multipronged methodological approach (see text box 1) first maps the AI-enabled value opportunities for automotive OEMs along the three application areas of process, driver/vehicle features, and mobility services (Chapter 1) The second chapter breaks down and quantifies these opportunities in terms of process-, driver/ vehicle- and mobility-services-related opportunities Finally, the third chapter outlines the strategic actions that OEMs should take to fully capture the AI-enabled value opportunities in both the short and long term
1 Cross-reference Smart moves required the road towards artificial intelligence in mobility (McKinsey, September 2017) and Smartening up with Artificial Intelligence (AI) What's in it for Germany and its Industrial Sector (McKinsey, April 2017)Artificial intelligence: automotive's new value-creating engine
Our analyses yielded the following key insights, which will be explained in more detail in the course of the report (for details on our sources and methodology see text box 1 and the appendix):In the short to medium term, there is a substantial industry-wide AI-enabled value opportunity, which by 2025 will reach a total accumulated value potential of around USD 215 billion for automotive OEMs worldwide This corresponds to the value of nine EBIT percentage points for the whole automotive industry, or to an additional average productivity increase of approximately 13 percent per year2 a significant value to boost the industry's regular ~2 percent annual productivity aspiration Most of this value is derived from the optimization of core processes along the value chainEven in the short term, AI can lead to efficiencies and cost savings across the entire value chain and can create additional revenues from vehicle sales and aftermarket sales Most of the value is generated through four core processes In procurement, supply chain management, and manufacturing, efficiencies lead to cost savings of USD 51 billion, USD 22 billion, and USD 61 billion respectively In marketing and sales, AI-based efficiencies both reduce cost and generate revenue, leading to a total value potential of USD 31 billion for this processWhile AI-enabled driver vehicle features and mobility services can generate substantial industry-wide value in the long term, these features and services only create limited value on the industry level in the short term Nevertheless, generating value from these features and services is important as individual OEMs that outperform competitors with their driver/ vehicle features and mobility services can gain substantial market share These gains in market share by technology leaders are, however, small compared to the risk of losing a significant part of the customer base for OEMs that are falling behind on these featuresFour key success factors enable OEMs to prepare for the AI transformation and to capture value from AI in the short term: collecting and synchronizing data from different sources; setting up a partner ecosystem; establishing an AI operating system; and building up core AI capabilities and a core AI team to drive the required transformationOEMs need to begin their AI transformations now by implementing pilots to gain knowledge and capture short-term value They should then establish the AI core to develop an integrated view on AI across the organization Finally, this will enable OEMs to scale up and roll out an end-to-end AI transformation to systematically capture the full value potential from AI and build up capabilities for their long-term ACES strategies
2 While this value is generated around the automotive OEMs' business, not all of this value can be exclusively captured by OEMs because other players, including suppliers, system integrators, and tech players, will try to capture some share of the value Fierce competition between automotive OEMs may also result in some of the value being passed on to customers In addition, some investments are required for the initial implementation of AI use cases and some (comparably low) costs arise for the operation of machine learning (ML) Nevertheless, we expect that automotive OEMs can capture the largest share of the potential value Artificial intelligence: automotive's new value-creating engineText box 1: How we derived insights sources and methodology
Our main sources includeThe McKinsey Auto 2030 market model, which is based on scenario-tested development of the disruptive ACES trendsMore than 100 discussions with AI experts, mobility executives, and functional experts on areas including manufacturing, supply chain management, sales and marketing, and ITRelevant market reports on digital disruptions, AI, and automation as well as annual reports from all major automotive OEMsMore than 15 analyses on specific industry perspectives, eg, how OEMs are investing their resources and what margins can be achieved How we derived the value potential We developed a use-case landscape along the entire value chain and quantified all major use cases by identifying the status quo for a typical OEM as well as the target state for full AI application Estimating differences in costs or revenues and margins then yielded the value potential (further details on the methodology used are provided in the appendix) The following is an example for the important overarching manufacturing use case, in-line quality control, which is relevant for stamping, body shop, paint shop, powertrain, and final assembly (with some variations between these manufacturing steps):
Status quo at a typical OEMQuality control is partly carried out by machines, partly manuallyManual quality control often has a low detection rate for smaller issuesIf a quality issue is detected, a manual intervention is requiredLimited learning can be taken from issues detected as there are multiple interdependent parameters and operators typically do not know how or which parameters to optimize (making changes is too risky) Target state required for fully capturing the value opportunityContinuous in-line quality control for automatic detection and high-accuracy detection of quality issues, for example by recording video, sound, and process parametersImproved manufacturing processes based on the incorporation of feedback from the detection of issues affecting quality, and therefore a large reduction in quality issues Why AI is required: AI is required for analyzing previously unavailable or indecipherable data (eg, video or sound which previously could only be interpreted by humans), in order to detect quality issues AI also has the ability to help detection and analysis mechanisms, improve its own accuracy by continuously learning from the issues detected, and optimize manufacturing processes by incorporating feedback and adjusting the control parameters accordinglyArtificial intelligence: automotive's new value-creating engine
EXAMPLE QUALITY CONTROL USE CASE。