The automotive industry today is vibrant and dynamic, and is often seen as a proxy for the economic and manufacturing prowess of world leading economies. Significant upheavals – both technological and managerial – have constantly led to better industrial throughput and performance. The automotive industry is not only regulated and cost-sensitive, but also increasingly at the intersection of multiple emerging technologies. In a series of blog posts, we discuss the impact that data will have in the automotive industry in the future.
General Industry Features
As of 2016, the vast majority of industrial production within the automotive space is consolidated. While ancillary suppliers provide key technological and product development capabilities to Original Equipment Manufacturers (OEMs), the system-level innovation and product level innovation that leads to innovative new products generally come from a few big names, that are design houses, well known automotive manufacturing brands and conglomerates that own multiple brands.
The OEMs in the automotive industry depend significantly on the supplier value-chains for cost competitive products, and most of these are serial/mass production houses. Emerging markets such as India and China have the greatest promise at the moment, both in terms of production volumes and in terms of their being new markets that are receptive to new business models and new value propositions in terms of transportation. The vast majority of young, informed consumers in these two markets (and similar markets) are faced with busy, chaotic and crowded urban centers and less of a need to own a vehicle, and a greater need to access a vehicle instead.
Data Driven Management in the Automotive Industry
Management styles have emphasized data driven decision making in the automotive industry more than in many other industries. For decades, quality management movements have influenced decision making on the manufacturing shop floor. There is a strong culture of quality and a focus on data driven quality approaches in many of these companies. Approaches like Six Sigma and approaches based on the Toyota Production System have been institutionalized in many automotive manufacturing companies and such organizations often have extensive data sampling, collection and hypothesis validation processes in place, for quality data and for other data analysis. Warranty analysis and warranty cost rationalization is another key area within the automotive industry where statistical engineering principles find use. In this sense, the automotive industry is uniquely positioned to gain from the revolutions in digital data collection, storage and representation and analytics.
Technology and Disruption in the Automotive Industry
While some companies have embraced emerging technologies such as advanced displays, wireless communication systems and additive manufacturing in order to compete, most are focused on a few key technologies that are central to the value proposition of the vehicle itself. Business models have changed in the last decade or so, with access economies enabling vehicle use as a service, thanks to taxi aggregation applications. In some ways, this has refocused companies on the key value the customers gain from the vehicle. Vehicles will increasingly be seen less as a necessity for urban life and more as an enabler of a certain living standard. The advent of self-driving cars pioneered by technology driven companies such as Google and Tesla have reinstated interest in stagnated sectors of the automotive industry that now see themselves as ripe for disruption.
The rise of tech startup unicorns has affected large technology companies in interesting ways. Many of them have had to work around new and interesting problems and use their scale to offset the risk of being out-competed or disrupted. In the automotive industry, which is more capital intensive and where the duration from thought to thing is longer, there is less risk of this. However, the availability of hardware and systems to become technologically convergent has affected the automotive OEMs’ and their way of thinking through key automotive systems. Vehicle-bound systems are now being migrated to distributed systems, ranging from music and media players to global positioning systems and radars. Key vehicle systems are now able to communicate and be in concert with one another, and the data from them can easily be connected. There’s excitement in the automotive community around the incorporation of advanced computing power right into cars themselves, and it is expected that CPUs with significant computing power will power the processes and systems of future vehicles.
Automotive Industry and the IOT revolution
It is in technological context that the Internet of Things revolution is gaining ground in the automotive industry of today. The IOT revolution is frequently associated with self-driving vehicles, or autonomous vehicles, and rightly so. However, the comparisons and possibilities don’t end there. There is a significant glut of IOT use cases within the automotive industry, starting from the collection of passenger and vehicle data from sensors, to the use of machine learning in real time to allow our vehicles to learn about us, and provide improved user experiences. It is increasingly clear that digital displays and interfaces of one form or another are integral to the way humans and computers will interact. The automobile is soon becoming an interface between an intelligent machine and a human, and in line with this development, there are opportunities to aid human control, with blind spot sensors, radars, eye tracking and driver drowsiness prevention technologies.
On the factory side of the industry, IOT brings many benefits to this already data-intensive part of the industry. The nature of industrial manufacturing processes is increasingly automated and many manufacturing plants are built to be extensively robotized.The use of IOT will allow companies with high levels of automation investment to make the most of this investment. Being able to collect data in real time at high sampling rates from machines and robotic processes will enable manufacturing engineers with lots of process optimization possibilities. IOT approaches can also be implemented in processes that still require human intervention, in warehouses, in part tracking and supply chains, and related areas of the automotive value chain, to leverage data driven decision making. Already, handling the parts, warranty and parts shipment in a typical automotive company is a complex value stream, which requires many departments to work in concert. It is these kinds of hard problems that require multidimensional data from various sources, that are easily solved by the use of big data technologies and data streaming and near-real-time data technologies. Near-real-time and real time expert systems can also break through processes that are currently heavily dependent on human expertise, such as paint quality, part quality and complex tests.
Problem solving and quality management are two more areas where automotive manufacturers can benefit from the large scale processing of data and machine learning. Already, approaches such as Six Sigma and Lean make extensive use of data collection but using sampling approaches. Using high frequency sampling and large samples of data for the analysis generally helps us uncover diverse patterns that may be beneficial in new ways to problem solving teams. Supplier quality teams and OEM quality teams alike can benefit from such advanced problem solving. This may even open up new business models and new value streams. A trend away from vehicle service as an add on to a vehicle’s purchase, towards vehicle service as an integrated part of the offering may be one such possibility. This possibility can be enabled by advanced predictive maintenance, which comes from a deep, data driven understanding of the various aspects of vehicle systems, subsystems and components and applying machine learning to reliability, durability and related problems. Data analysis could help in related areas, such as failure mode classification and avoidance, automated failure reporting or alerts, etc.
Improved energy efficiency is one more of the needs of the automobiles of the future. Product design and development teams will find use for advanced machine learning and decision support systems that can be built into vehicles, to not only design cars better, but also help the machines make real time decisions to improve the fuel and energy efficiency of vehicles. Whether fuel or electricity driven, the vehicles of the future can benefit from advanced algorithms built right into the systems that control the transmission and engine systems and subsystems.
At the sales interface, machine learning, AI and data driven decision algorithms can help identify and retain the best customers, and find other potential customers who may like the design, value proposition and the features of a company’s product. Recommender systems, that are making waves in the e-commerce industry, could be used in bold new ways to reach out to more customers than before. Customer churn analysis and customer loyalty analysis, targeted rewards and cross-selling are more easily done than before, and will continue to become part of decision support systems for marketing and targeted selling.
The automotive industry’s many challenges can be offset at least partially by the deployment of data-based decision making, and expert systems that rely on advanced data analytics to deliver value to the manufacturers and results. Customer delight of new kinds is now within reach of the automotive design and manufacturing teams in this consolidated industry. New challenges of upselling, cross selling and maintaining high quality in products and services are all enabled by using advanced analytics. Increasingly, the decision for automotive manufacturers seeking to improve sales, quality, profitability and customer and employee delight, is not whether they choose analytics and data driven approaches such as advanced machine learning and IOT, but when they choose it.