CSO at imec, a leading and innovative center in the active world in the fields of nanoelectronics and virtual technologies.
Studies through the U.S. National Highway Traffic Safety Administration (NHTSA) show that human error (e.g. error of judgment, speeding, fatigue, impaired or distracted driving, etc.) is the cause of 94-96% of all traffic accidents.
These staggering figures illustrate how difficult it will be to achieve governments’ ambitious road protection goals, such as the EU’s Vision Zero initiative, which aims to achieve a near-zero mortality rate on EU roads until 2050. It is not unexpected that the threat of human (driving) errors is minimized or prevented is one of the key price proposals for the arrival of self-driving cars (also known as self-employed or driverless).
Road tests with self-driving cars (VA) have been underway for over a decade. Today, car manufacturers claim to have driven millions of kilometres of autonomous driving on public roads. However, despite the ongoing buzz around the VA, it will take years of progression and thorough testing before self-driving cars are a component of everyday life and traffic.
As impressive as today’s tests are, they are sometimes carried out in ideal cases (i.e. in fenced or highly controlled environments that have been fully mapped in 3D). But as we all know, traffic in urban centers is anything still predictable: cyclists operating on the streets in a very artistic way, pedestrians who climb and go down the sidewalks, and motorists replace lanes, turn or avoid unexpectedly. Current driverless vehicle technologies are not yet in a position to meet these challenges.
Does this mean that VA’s promise is absolutely out of success for the foreseeable future? Certainly not. Researchers and (automotive) corporations around the world are making great strides to take the underlying technologies to the next level, making them more powerful and physically reliable. The progression of autonomous systems that cover all possible conditions and use instances with 99.9999% reliability only requires more time.
Technological innovation requires use instances and out-of-the-box models
In the end, the innovation effort mentioned above has a burden, literally. Today, for example, as a springboard to the arrival of vehicles without force in fact, brands have begun equipping cars with complex drive force assistance systems (ADAS). ADAS has the ability to avoid or mitigate the severity of automotive injuries with features such as adaptive cruise control, lane keeping assistance, blind spot warning, pedestrian detection and more.
However, the addition of these characteristics makes cars increasingly loved. Needless to say, the arrival of VA will mean even higher prices, which can make driverless cars too expensive for widespread short-term adoption.
So, despite the promise of technological advances, is the fully autonomous car doomed to fail even before it is introduced? Maybe not, if this comes with new usage instances and strategic business models.
An autonomous, on-demand round-trip service in Singapore
As mentioned, the existing generation is able to adapt to autonomous driving in low-speed, low-traffic or fenced-in environments. Consider on-site airport transfers or transportation passengers to and from the city’s suburban car parks and shopping.
In Singapore, an autonomous service was tested on demand back and forth on Sentosa Island last year, with 4 self-cellers traveling in a predefined 3.5-mile address. To use the loose service, consumers simply hailed a separate round trip through a mobile app or compromised kiosks near each stop. Each VA was “equipped with radar, lidar, GPS, odometry and computer vision to help stumble upon its environment and identify proper navigation directions, pitfalls and signaling.”
Initiated through local municipalities, audiovisual deployments on the outskirts of the city, combined with small-scale urban mobility responses and classic public transport characteristics, can be particularly the pressure of local traffic and parking.
Private cars: the end of an era?
But this is just one example of how VA can revolutionize mobility. Another example is replacing personal cars with subscription-based mobility-as-a-service (MaaS).
Since the price of ADAS-equipped and, in a later stage, autonomous vehicles will continue to increase, we should ask ourselves whether consumers will be prepared and have the financial means to spend several thousands of dollars extra on driverless cars that will stand idle in people’s driveways for approximately 95% of their lifetime.
In the future, any investment in more-stocked (and more expensive) cars can be expected to be offset by higher use. Enter the car exchange, where the right type of car (self-driving) can take others where and when they want transportation.
The key to success: converting the equation of the structure and costs of the automobile
In short, it can be said that a successful arrival of VA is very likely to go hand in hand with cutting-edge use cases and business models that extend the investment in a highly complex and beloved generation to many subscribers. As such, the position of innovation would no longer be prohibitive in the deployment of generation.
Of course, for this to work, consumers and the vast ecosystem of automotive players will have to replace their minds. Consumers will have to abandon the concept of long-standing personal cars, while the automotive industry will need to focus even more on delivering the most productive features and facilities imaginable rather than focusing on mass production and sales.