With over 100 million lines of code, the software fueling modern cars makes them the most complex pieces of technology that we use in daily life. It is also software that offers most room for improvement, and hence most room for differentiation: in the end, there are only so many extra cupholders and heated seats a carmaker can add to the manufactured hardware. The true value for consumers can be unlocked by integrating cars into the internet of things, increasing collaboration of hardware sensors for safety and fuel efficiency purposes, or improving the in-vehicle infotainment systems.
The argument for crowdsourcing in the automotive industry: value creation
Car enthusiasts have been tweaking their machines for decades: horsepower, turbo capability, hydraulics, looks. Some brilliant tweaks have since been adopted by the industry as best practices, fueling innovation in the hardware design. That alone should be argument enough to democratize software and allow the able crowd to decide development direction for the brains of a car. General public would contribute far-fetched consumer whims while app designers and passionate coders would bring them into life.
Another argument lies within: automakers simply do not have the capability to innovate in the space that lies far outside their core competency of assembling cars. Most recent attempts to steal real talent away from Silicon Valley have failed as top coders are rarely attracted to large, rigid organizations. Opening the source code and crowdsourcing improvement ideas could solve the talent problem.
Finally, recent events brought attention to another argument: crowdsourcing as control mechanism. In the aftermath of Volkswagen emissions scandal voices demanding access to automakers’ code are getting louder, claiming that the cover up would have been avoided altogether if car owners were allowed to enter the brains of the car. Even if most people would choose to stay out of it, more than enough out of the 1.2 billion car owners are savvy computer wizards to keep automakers in check.
The early signs of crowdsourcing in non-crucial vehicle subsystems: value capture
The idea of crowdsourcing in automotive industry is not exactly novel. Local Motors is an example of a crowdsourcing car design company. It engages a community of over 30,000 passionate and capable members to co-design car hardware from start to finish.
Open-source software is also starting to make its way into cars beginning with noncritical systems such as in-vehicle infotainment. For example, Ford is the first US automaker to pledge its proprietary code to an open-source alliance allowing independent developers to build apps on traditional platforms (iOS, Android) and integrate them directly with Ford’s vehicles.
One of the largest alliances in the space is currently being tested by the Linux Foundation: Automotive Grade is a workgroup bringing together automakers (e.g., Toyota, Nissan, Jaguar, Honda), chip makers (e.g., Intel) and mobile device makers (e.g., Samsung, Sony) dedicated to creating open source software solutions for automotive applications. This is a significant move towards openness given the unprecedented cooperation between ardent competitors; a sign that automakers see the large value capture potential in crowdsourcing.
My hypothesis on how the value capture in this space will play out goes back to the theory on commoditization: wherever one side of an industry becomes commoditized (in this case, car hardware), the profit pool shifts towards a critical but not-yet-good-enough subcomponent (in this case, car software). I suspect this will happen here: in order to keep high level of profitability, the carmakers will HAVE TO step up their game in the software space, starting with the non-critical systems, but quickly moving to the real brains of a vehicle. Moreover, it will be the few first movers whose systems will win. And harnessing the collective wisdom of a vast number of consumers is the surest way of guaranteeing that victory.
The challenges ahead
The challenges are many. The first question that comes to mind – will the crowd want to participate – is the easiest to address. History of car tweaking within automotive shows that people like to have an individual say in performance of their machines and don’t shy away from their perceived complexity. Current landscape is also full of examples of (currently semi-legal) contributions to software: a passionate Tesla driver recently tapped into Tesla’s operating system and created an app called Visible Tesla allowing its users to deploy commands that remotely control the car. Upon finding out, Tesla didn’t even mind, possibly recognizing that herein lies the future.
A much larger challenge lies within the legal realm: current legislation passed initially to protect DVD content, the Digital Millennium Copyright Act, restricts the circumvention of proprietary digital mechanisms, by now also in cars. The reasons are indeed compelling: the flip side of allowing crowds to write car code is that some will do so with malintent, deploying viruses that override safety systems. This will mean large investments on the part of the carmakers to weed out the hackers.
The final challenge has to do with exactly that: investments. Integrating crowd’s ideas and code given the liability and security issues, testing requirements, and the complexity of systems’ integration will undoubtedly be expensive. However, it may turn out that not bearing that investment will lead the carmakers to face the ultimate cost: being run out of business by competition that did not fail to innovate.