Software is Reinventing the Wheel

The automobile is undergoing a fundamental transformation. Great innovations are already on the horizon. At the heart of each lies software. As cars are increasingly defined by their software, what effects will this have on the automotive industry?

In this article, we discuss how software is revolutionizing the concept of vehicles:

• The 'Why'
• Factors
• Roadblocks 
• The Path Ahead

Why Vehicles Will Be Defined by Software

Cars have gradually gone from pieces of electromechanical hardware to computers on wheels. As such, software is becoming the primary source of value in vehicles. Appropriately, software development is consuming an ever larger share of carmakers' R&D budgets and by 2030, software is expected to account for 50% of vehicles' bill of material (BOM). The ballooning investments by automakers in software development reflect a recognition that in the near future not only will software drive the majority of value creation, but it will fundamentally shift the business model of the industry. Automobile manufacturers will compete and seek to differentiate their brands on the basis of software. Beyond changing their unique value propositions, software capabilities will fundamentally change the selling model–from one in which the original equipment manufacturers (OEM) have one-off sales to one in which software and service premiums provide continuous revenue streams. 

Factors Driving Software-Centricity

The newest innovations awaiting the world will require highly advanced and intricate software. Autonomous driving makes this point the strongest. It will require artificial intelligence to identify objects as humans or cars. It will require high degree of connectivity with other vehicles and the world around it. And it will need to do these and countless other calculations instantaneously or risk injury, while also maintaining incredibly impervious cyber security. Autonomous driving will require large quantities of hardware devices, which in turn will require even more code. Modern luxury vehicles already have hundreds of millions of lines of code. This will only skyrocket as vehicles become autonomous, electric, and interconnected.

Another factor driving a software-centric business model for automakers is the so-called hardware commercialization. Hardware commercialization denotes the increasing homogeneity of car parts. As the differentiation in vehicle hardware decreases, so do the profits as well. And when parts manufacturers can't compete on the basis of the cars' parts, they need another differentiating factor to drive profits–software. Software services are already gradually becoming an important stream of revenue. An example of the shift in the business model can already be seen with Tesla, who offers an FOTA (firmware-over-the-air) software upgrade to customers for $100/month. Experts expect the subscription business model to proliferate as it provides continuous, sustainable cash flow.

Roadblocks Impede the Future

As the cars grow more intelligent, the number of electronic control units (ECU) explodes. In current vehicles, these ECUs include everything from the engine operations, powertrain modules, body control modules, etc.. As vehicles gain new functions, they of course also need more ECUs and with additional ECUs, wiring becomes both more complex and expensive. There are also limits to the microcontroller units (MCU) that serve as the basis of ECUs. In order to reach the fourth level of autonomous driving (L4), it's expected that vehicles will need a computing power of 100 tera operations per second–ten times as many operations/second as what's needed for level two autonomous vehicles. All of this will require additional changes to the electric & electronic architecture. This is further complicated by the fact that ECUs are often supplied by different suppliers who then use different coding languages, which makes ECU software inefficient and redundant. It is likely we will see the computers within vehicles become more centralized.

Beyond the hardware challenges, the automotive software development organization is in desperate need of an agile transformation. The top-down traditional model of software development is an antiquated relic of when the end of the value cycle was the sale of the vehicle. It is too cumbersome and cost intensive to deliver the rapid software iterations needed for a software service and subscription model. Vehicle manufacturers will need to reduce the time-to-market and ensure that software services are developed with the end user in mind. An agile approach will need to be adopted. But agile methodology brings with it challenges of its own, especially in an organization as vast and complex that of a car manufacturer. It risks silos. Cross-functional collaborative teams can mitigate this risk and promote efficiency. Software analytics can also smooth out an agile transformation by creating deep transparency into the software development, allowing company leadership to track the progress of various digital initiatives and highlighting areas of concern for managers to intervene.

The challenges and appropriate strategies for members of the automotive industry vary widely depending on their positioning as a luxury or mass-produced brand, if they continue to produce combustion engines or switch to electric, and if they are the original equipment manufacturers or parts manufacturers. Many vehicle producers, for example, are having to develop their own operating system (OS) or are partnering with technology firms to various degrees to do so. The transformation of the industry has also lent newfound importance to players in automotive, such as chip manufacturers or algorithm developers. 

The Path Ahead

The path ahead varies greatly depending on the company's role in the supply chain. But regardless of if it's the car manufacturer, a parts supplier, or a pure play software provider, they do all face a similar challenge in that software is reinventing the car, but that its development remains unmanageable and complex. For many players in the automotive industry, especially incumbent carmakers, they lack the software know-how and talent required to surmount the challenges ahead. Digital boardrooms offer one part of the solution by providing a level of transparency into the software organization that enables the strategic management and planning by company leadership.