Hype check: Why self-driving cars’ long-term impact is underestimated

Paul Asel 2012 11

Connected car makers are getting a reality check. Pilot programs for fully and semi-autonomous self-driving cars have ignited a hype cycle that’s turning in on itself, with even well-funded initiatives from Apple Inc. dialing down expectations.

Perhaps interested parties, from Tesla Motors Inc. to Ford Motor Co., are coming to terms with the real-world challenges of revolutionizing all things familiar and known about transportation. Fortunately mankind has a history of technological advancements, lighting the paths to mainstream adoption.

Today, one venture capitalist is tapping his nearly 30 years of experience in tech to become a voice of reason within an industry anxious to fulfill futurist dreams. In an interview, Nokia Growth Partners managing partner Paul Asel (above) tells SiliconANGLE why the market is overestimating the short-term possibilities of autonomous cars and underestimating the long-term prospects. His company’s global team provides unique perspective on technology trends, alongside strong analytics. The result is confident investments and lucrative exits, including the sale of UCweb by Alibaba Group Holding Ltd. for $4.7 billion. Nokia’s investment arm has allocated $100 million to the connected car movement, with another $350 million for the broader IoT sector. As the group’s resident connected car expert, Asel envisions the ripple effect of self-driving cars, and outlines three barriers to mainstream adoption.

Q: What’s been your historical view of tech hype cycles, and how does this experience translate to the connected car market?

A: Having been an investor for approaching 30 years and having seen many of these cycles, my observation is we overestimate impact and opportunity in near term, and underestimate the opportunity in long term.

Looking at 2G vs 3G vs 4G [wireless telephone technology generations], adoption cycles took substantially longer than expected in each case. In 1999 3G was much hyped and many were investing based on rapid conversion to the mobile internet. I remember talking with an advisor at the time who had been in the venture business for 40 years.  He asked, “When do you think 3G will reach critical mass?”  I responded with the longest time frame I thought to be credible and predicted the year 2002.  It wasn’t until 2007 that we saw 3G really take off, and mobile-first solutions emerged after 2011.  The industry and I were off by a factor of four.  And yet, few would have imagined cell phones playing the role in our daily lives that they do today or the disruption the mobile-first companies are causing across numerous industries.

We are highly committed to the connected car space. We have a $100 million auto tech fund, investing in nearly a dozen companies globally, including Peloton Technology and CityMaps in the United States; Drivy, Deliveroo and Moovit in Europe; and Zoomcar in Asia. Each of these companies leverage new auto technology to change the way we think about transportation and mobility without relying on the promise of autonomous driving.

We think autonomous driving will have an enormous impact on all industries that involve transportation and mobility as part of their business. Yet, we also think adoption of autonomous driving will happen incrementally over decades rather than years as many people anticipate.  An example that goes back to the telecom industry is the Nokia Communicator, which came out in the early 1990s. It had much of the capabilities that came out in early versions of Apple. But it took another eight to 10 years for supporting technology to bring the experience to life and to gain sufficient consumer adoption for an ecosystem to develop that enabled many of the services that we value so much today. So it has been a decades long process even with a simple device like a phone.  For more complex machines where safety and more infrastructure investment is involved, the adoption process will be considerably longer.

Q: Which factors for self-driving cars’ extended time cycle should autonomous driving advocates NOT ignore?

A: There are at least three barriers to adoption beyond perfection of the technology, which itself will take 5-10 years to be fully implemented. Look at the SAE levels of autonomous driving. What’s being implemented today is partial and conditional autonomy. What’s being professed and promised is full autonomy. We’re still at least five to 10 years away before the technical requirements of full autonomy are met. This includes challenges we understand today – sensor fusion, integration into the critical car functionality, technology that works in all road conditions, and navigation systems to work on all roads in all contexts – as well as those we will confront over time as the technology progresses. There is a lot that still needs to happen on the tech side.

Beyond tech hurdles that need to be addressed, there are three additional barriers to widespread adoption:

1)  Automotive design cycle – Autos currently have a five-year design cycle. The industry is moving fast but relative to its own timeline. Also consider that traditional automakers really do not want full autonomy to happen. Full autonomy will actually shrink the number of traditional cars on the road by at least 50% as fewer cars will be required to serve our current needs.  Automakers are driven by the threat of obsolescence as much as new opportunity.

Early adoption of autonomous technology will be driven by those who have the most interest in the benefits of autonomy.  New entrants such as Tesla are highly interested in autonomous driving because it’s an opportunity for them to disrupt the incumbents. Autonomous technology is of great interest to any firm with costs associated to drivers.  Taxis, rideshare programs, fleets, cargo and shipping industries all benefit from autonomy through lower labor costs.

2)  Regulation – The regulatory approval process, which could take decades, is even longer than the automotive design cycle.  We have just seen the release of guidelines last month on autonomous driving.  Yet there are concentrated risks and distributed benefits to autonomous driving.  While consumers may benefit, risks are concentrated with the incumbent automakers.  The regulatory process will be slowed by lobby groups, and consumers and regulators are unlikely to be strong proponents until safety concerns are fully addressed.

The drone industry is instructive as an example.  Commercial drone regulation has been delayed for more than five years over concerns relating to privacy and safety.  We are seeing commercial experimentation in Africa as a first mover because the benefits of drone delivery are higher there as a result of poor infrastructure and inadequate alternatives.  Some of the first implementations of autonomous driving have been in emerging markets as well, for instance, with the mining industry in South America.  But while the regulatory environment will be less of an impediment in emerging markets, autonomous driving implementations will be limited by poor infrastructure and lack of high-definition maps needed for navigation.

3)  User behavior – Often we find that user behavior is the biggest barrier to new technology adoption.  Consumers will be big beneficiaries of autonomous driving, yet adoption could well be generational.  User behavior is resistant to change, and many of us simply love driving cars.  I see user adoption playing out across four different levels:

  • Special use cases – universities, parking and other off-road situations where convenience is high and risks are low
  • Where economic merits are high – long haul trucking, taxis and ride sharing, local deliveries
  • City dwellers that have a high cost of ownership for cars
  • Consumers at large

Q: Can we chalk up some connected car advancements to self-fulfilling prophecies? Don’t we need a bit of hype to incite innovation?

A:  Many have an interest in lauding the potential of innovation.  Naysaying is not newsworthy while one-upping others’ predictions is highly quotable.  This is why the Gartner hype cycle is so pervasive, and one reason we have a tendency towards boom/bust cycles in stock markets. Flights of fancy are fun, whereas fits of depression are best deferred as long as possible.

But where there’s smoke there’s fire. And this is where the second part of the paradigm come in.  It is easy to overestimate short-term impacts but hard to think through the full range of long-term ramifications of technology disruption. So our limited imaginations cause us to underestimate the long-term opportunities and consequences of new technologies.

One book that illustrates this well is the book “How We Got to Now” by Stephen Johnson, which discusses six innovations that made the modern world. What he did remarkably well was highlight the root cause of an invention and tracesthebroad ramifications of inventions in related industries and across society.  Alfred Chandler also traces causal links wonderfully in “Inventing the Electronic Century.”  They’ve informed, to a large extent, the diffusion of innovation.  Both illustrate the point that the more we look back the further we can look forward.

Q: What are the differentiated ways autonomous cars will impact the market moving forward?

A: One is the changing of cities. We talk about the diffusion of innovation – one image I typically reference is a city as it appeared in the early 1800s, the 1950s and today. The advent of industry and adoption of different modes of transportation have had tremendous implications on the design of cities. In the next 50 years, cities will be redesigned.  Cities will serve as denser gathering points, while residences will be more distributed. The average person spends 50 minutes in the car daily. With autonomous driving, the effective cost of commuting is lower, and people can live in a more distributed environment with communities sorted according to interests and preferences.

How will it ripple through to other industries? Insurance will be significantly impacted. Over 90 percent of accidents are caused by human error. The insurance market as we know it today could shrink by 50 percent or more in the next 50 years or will be substantially altered. We have invested in Zubie, a dongle that goes into the car.  But we are not investing in a dongle. Rather, we are investing in a company that perfects predictive maintenance and usage-based car insurance. As people become more aware of risky driving behavior, we have found that driver scores improve by as much as 25%, which should reduce accident rates even without autonomous driving.

The cargo industry and transportation of goods and services will also change dramatically. When the cost of transport is lowered with autonomous driving, we will think about trade and where goods and services are located in a different way. We’re actively looking at trade and transport as areas of investment, working with Deliveroo for short-haul services. We are looking at the long-haul space as well.