An Operating System for Self-Driving Cars
Updated: Jan 31
As the year 2019 comes to a close, I am reflecting on what we are building here at Apex.AI and how it impacts the mobility industry.
Looking back ten years, the automotive industry was in a transformative epoch. In 2009, Google launched its self-driving car project, seeding the race of Silicon Valley tech companies into autonomous mobility, while the largest US automakers were bailed out from the verge of bankruptcy with $80 billion from the federal government.
The smartphone industry was in an equally transformative phase. Nokia was the global mobile phone market leader in 2009, Apple’s iPhone was two years old and Samsung launched its first Android phone. Barack Obama was sworn into office that year. What a difference ten years can make!
The cumulative effect of small advances can be surprisingly large over a longer time-frame like ten years - an effect sometimes referred to as Amara’s Law.
The lessons in corporate strategy have been equally surprising for technology companies.
Ten years ago, there were many Operating Systems (OS) and software development frameworks that were in consideration as alternatives to iOS from Apple. Some were phone OEM’s homegrown efforts, and some came from OEM consortia. Today, open-source Android has won, and accounts for about 75% of the global mobile OS share, with Apple’s iOS accounting for the remaining.
A qualitative study from researchers at Aalto University and INSEAD (2015) (1) shows how OS choices directly led to Nokia’s fate in smartphones.
“Apple’s iPhone was a major discontinuity for the whole industry: for the first time, differentiation lay in software rather than radio technology. Nokia, hoping to ensure sustainable profit, decided to continue improving its proprietary OS, Symbian, which
both internal and external developers generally described as difficult to work
with. [...] Consumers’ responses to Nokia’s phones became increasingly negative.
[Nokia engineers knew] they needed a better OS to match iOS, they also knew
that developing it would take several years. But they were afraid that publicly
acknowledging Symbian’s inferiority to iOS would torpedo sales figures and
shatter internal morale in the short term.”
The situation in the self-driving car industry today is remarkably similar to the smartphone industry a decade ago. Currently, more than one automaker is trying to develop an OS and development framework for the hugely complex autonomous driving stack. Jan Becker, co-founder and CEO of Apex.AI, has written about why the autonomous mobility technology stack should soon converge to a common software framework, with a standardized and open architecture. A modern software framework allows hardware and applications to innovate independently and more rapidly. The consequent explosion of technology and innovative uses in autonomy is yet to happen, but the building blocks exist today.
Looking back a decade from now, we might see the coming decade as remarkable for autonomous driving as the previous decade was for smartphones.
We are building Apex.OS (2), a software development and deployment framework that automotive and autonomous mobility companies are adopting. The open source framework on which it is based, ROS, just had its largest developer conference ever, attracting notable sponsors like Toyota, Microsoft, AWS and Bosch. For the first time, a critical mass of users and a common need for a software framework for self-driving cars have come together.
Not every automaker today will survive the next decade. But it’s safe to say that among those who succeed, the common thread is likely to be an open, standards-based software framework leading to faster and better innovation across the mobility stack. The timeline of self-driving cars depends on the choices the industry makes.
(1) - “How Nokia Lost the Smartphone Battle”, Administrative Science Quarterly, 2015 (link)
(2) - Apex.OS is based on ROS, a open-source meta-operating system for multi-sensor multi-actuator systems (http://wiki.ros.org/ROS/Introduction)