Will the self-driving car industry remain a perpetual hype?
The word ‘Hype’ as defined in Merriam Webster Dictionary: “Stimulate”, “Enliven—usually used with "uphyping”, “Increase”, “Put on”, “Deceive”, “To promote or publicize extravagantly”, “Publicity especially promotional publicity of an extravagant or contrived kind”.
Understandably, every new big idea or vision of the future requires some level of hype. Big ideas are rooted in promised benefits that are transformational in nature. The path to progress is based on transforming the present into a better future, whether the promised benefits are economic, social, environmental or financial. A little hype is a necessary tinder to light the fire - it is required because it creates excitement, enthusiasm and attention that leads to attracting investment, talent, venturing, skill development and knowledge creation, with the hope that from increased activity and momentum, the reality of the idea will emerge to deliver the promised benefits.
Most if not all big idea hype is a not a lasting affair and usually within 2-3 years, either the idea falls by the way side, and is consigned to the heap of failures, or starts to deliver some real outcomes. Gartner’s hype cycle exemplifies this on an annual basis. We saw this in the dotcom boom when from the carnage of early 2000’s dotcom bust, emerged commercial powerhouses like Amazon and ebay that defined the age of e-commerce. We saw the hype of Internet of Things (IoT) when in 2010, major large corporates were predicting 50 billion connected devices by 2020 with many hype-driven slogans, but then within 3 years we saw something more real with the emergence of smart watches/activity monitors, smart thermostats, connected door bells, home assistants and the like.
The one factor that distinguishes the current hype of self-driving car technology from every other hype in the past is that policy makers and government agencies seem not only caught up in it, but appear to have chosen to become active participants. Automated transportation is seen as a strategic industry and the underlying technology has assumed a security dimension linked to economic domination of the future. Self-driving technology has the potential to disrupt the automotive manufacturing industry and transportation services. Between the two, we are talking about 10’s of trillions of dollars and 10’s of millions of jobs worldwide. With numbers like this, it is bound to be political, but politics should not be used as an excuse to produce or promote hype.
We have seen a few reality-checking news items beginning to trickle in since early 2019 about the prospects of self-driving technology. Investment activity that peaked in 2017 through 2018 has resulted in a handful of players (mostly in US and China) having amassed very large financial war chests – but not necessarily having the technology ammunition. Capital definitely creates enabling conditions for invention but cannot guarantee invention. Self-driving technology is still in the process of being invented, though we consistently hear otherwise. The hype is big, it is persisting, and now turning into ‘fake news’. This persistence of hype with a topping of fake news cannot be good for the industry because it keeps inflating expectations without delivering reality for longer and longer. The gap between the reality trickle and hype driven fake news explosion does a disservice to the emergence of real stuff by making it harder for incrementally better outcomes to be accepted in the interim.
It is high time this be called out – those who claim success by hype, must prove it
For context, let’s examine the recent self-driving technology demonstrations that can be considered on a range of transparency levels. The most transparent demonstrations usually show real-time in-cabin driving video alongside a clear view of the outside scene and showcase the entire journey from start to finish. Acting as a virtual drive, such demonstrations enable close scrutiny of the autonomous system’s performance. Seldom do companies feel confident enough in their technology capability to put out demonstrators like this, however, when such demonstrations are showcased, they definitely stand out.
The second type of demonstrations have video footage sped up several times real-time speed. These demonstrations seem visually more engaging and convey a more agile impression of the self-driving technology. However, such demonstrators allow cover up of faults, corrections and disengagements, through subtle video editing which is hard to notice in sped up footage. Slowing down the video to a real-time playing speed often reveals a different story.
The third category of demonstrations, which firmly belong in the grey area, showcase technology with no accompanying in-cabin footage while the system is ‘engaged’ or ‘driving’. Usually such demonstrators show only a video of the outside scene interspersed with some fleeting in-cabin shots. These provide no technology evidence as the vehicle could have very well been driven manually throughout. To build false credibility, such videos are often interspersed with a few seconds of technology visualisations, mixed in with snippets of autonomous driving and a little icon or text that reads “fully autonomous”, “100% autonomous” and the like. In these demonstrators, the most challenging scenarios are never showcased with in-cabin footage, and instead, a vehicle is shown merging into traffic or negotiating a roundabout with the skill of a human driver, and only an outside shot of the vehicle to serve as evidence.
Finally, the most hyped-up and incredulous demonstrators are claims with no evidence of automated driving at all. These types of demonstrators often appear in major PR media outlets accompanied by bold headlines. Usually, the only evidence provided to back up the claims of automated driving are a picture of a sensor-laden car or a snapshot of a safety driver with their hands hovering above the steering wheel. Typical hype-filled headlines follow a common theme, such as:
This top-secret car drove hundreds of miles autonomously from here to there…
Such and such project has broken the longest automated driving record…
This self-driving car learned to drive like a human in under an hour…
Driverless cars arrive at such and such city – get ready…
Most complex driverless trials launch in this city…
This is the worst type of hype as it contorts reality for the public at large through smoke and mirrors, and makes it harder for real innovation to break through. If we are to move forward as an industry, we must strive for transparency of technology, even if system performance is less than perfect. Embracing the shortcomings of a technology enables iterative improvement, which results in progress. The more everyone understands about the true state of technology readiness and the technological challenges yet to be overcome, the sooner the industry can move from R&D towards a plateau of productivity.