When will autonomous mobility really become a reality?
Autonomous mobility is like a young child star but too young for now
“From 2020 you will become a permanent backseat driver” was writing The Guardian in 2015. We are now in 2022 and robotaxis are as rare as flying cars. What happened to the announcements made by General Motors, Honda, Toyota, Tesla and Google? When will autonomous vehicles really be sprinkled among our life?
As of today, if you take a deeper look, you can find autonomous vehicles on the road but they are sparse. In the US, Waymo-Google and Cruise-GM launched experimentations in cities like Phoenix, San Francisco and Los Angeles. In Beijing, Baidu was granted an authorisation to deploy 100 vehicles in the upcoming months. How come the most advanced countries are only starting to switch from experimentation to commercial pilots whereas The Guardian was promising us to become a “permanent backseat driver” in 2020?
Autonomous mobility is not a mirage, it’s a breakthrough revolution that will infuse our life step-by-step
Self-driving technology is way more complex to be developed than expected. “Generalized self-driving is a hard problem, as it requires solving a large part of real-world AI. Didn’t expect it to be so hard, but the difficulty is obvious in retrospect. Nothing has more degrees of freedom than reality.” confessed Elon Musk in July 2021.
Indeed, imitating the learning process of humans that takes years to become a good driver involves numerous top-notch technologies. To make the right decision, at the right time, the car must be equipped with high-quality sensors, deep AI algorithms, state-of-the-art processors and chipsets. Another major roadblock to the large-scale deployment of such vehicles is social acceptance. Even if the technology is mature enough to be safer than human drivers, it will not be sufficient. The technology must have no margin error, it must be 99.9999% safe.
Taking these new elements into consideration, OECD’s International Transport Forum sets a targeting date for large-scale deployment that is much more realistic. They estimated that 25% to 40% of car journeys would be autonomous worldwide by 2050 which leaves us quite some time…
This being said, autonomous mobility will still be a breakthrough revolution for our societies. It will re-shape the industry, deeply change our day-to-day way of moving and living. By cutting down operating costs to near-zero, shared mobility services will see a significant boost. They will deploy their transport network to higher granularity and wider, meaning in dense and rural areas. Therefore, it will strongly increase their efficiency leveraging a higher frequency and a better flexibility of the services. Take a look at Goggo Network’s article on AMNs to read more about this topic (link).
This major paradigm shift, will not only generate economical benefits but also social and environmental improvements like safety increase, energy optimization, reduction of traffic, price decrease and a wider network coverage (especially in rural areas).
Even though massive investments were made, the technology needs some time to grow
The private sector has genuinely gotten the power of such a revolution. More than $28B has been invested into autonomous vehicles in the last 3 years. Led by the US and China, they are expected to keep on fast-growing in the upcoming years (cf. image 3).
We could wonder why these companies are investing so much whereas the technology needs some time to grow and scale-up. First, this market has strong barriers to entry, which means that high capital expenditures are required to develop the technology.
Secondly, AI chips and software are moving to the center of the conversation, as the marathon for autonomous vehicles charges onward. However, creating a powerful and complex kind of chip needed to train AI algorithms efficiently is expensive and challenging. Nontraditional chipmakers including Google, Amazon, Microsoft, and latest Tesla now design their own chips. Traditional chipmakers such as Nvidia expanded its automotive influence by offering DRIVE Infrastructure, end-to-end solutions for training, development, and validation of autonomous vehicles. It has, to a certain extent, lowered the barrier to enter the game and boost the proliferation of robotaxis and self-driving vehicles. For example, many Chinese automakers — including SAIC, NIO, IM, Xpeng, Li Auto — are deploying Nvidia technology. The major AV developers (Argo AI, Aurora, Zoox) and smaller ones (Nuro, Momenta, Navya) are also working with Nvidia. By April this year, Nvidia announced its automotive pipeline exceeded $8 billion for AI-based mobility solutions.
Thirdly, leaders of the industry pivoted their strategy from people transport services only to logistics services. For instance, Waymo-Google’s expanded its strategy from launching robotaxis only to developing autonomous trucks, last-mile delivery vehicles or joining projects such as CAVENUE that consists in building an AV-only lane on highways. This means new competition to existing players such as Nuro in the US. In China, big players such Alibaba’s Cainiao Network, JD.com, and Meituan also entered the competition to the existing players such as Neolix. These new initiatives are technologically more feasible thus, offering a shorter time-to-market and return-on-investment. In a nutshell, despite the substantial technology progress, the difficulties and complexity of achieving autonomous mobility was underestimated by the automotive industry.
This gap will be mitigated in the coming years with heavy investments into the field and increasing numbers of players joining the game. While turning everyone into a “permanent backseat driver” may still be the moon shot, new initiatives are digging the gold mine in the logistics services and other low-speed use cases.
Taking these use cases into account, we may broaden the concept of autonomous mobility and depict three phases of how this young child star could grow mature. We are happy to show you the pictures of these phases in the next article.
Stay tuned.
Octave Labbé, Project Lead @ Goggo Network
Jiayun Shen, Deputy Head Outpost China @ Swiss Post
Goggo Network was founded in 2018 by serial entrepreneur Martin Varsavsky and ex-McKinsey Associate Partner Yasmine Fage to lead the driverless revolution in order to solve the transportation challenges of today.
Swiss Post is the national postal service of Switzerland. A public company owned by the Swiss Confederation, it is the country’s second largest employer with about 54,000 employees. The group is based in Bern and has branches in 25 countries.
