Robots Moving People

My bottom line from the London Mobility Summit – Autonomy event is: enough with endless trials; we need courage and determination. Here is my take on the state of autonomous vehicles, I’d love to hear your thoughts. Let’s start, robots moving people: 

1. Stuck on the slope. The Gartner Hype Cycle aptly describes the current state of autonomy - the industry is slowly climbing out of the “Trough of Disillusionment,” with the emphasis being on slowly. Autonomy, once promised by 2024, now seems distant. Most use cases will struggle to achieve scale before 2030.

2. Safety, regulation and an incomplete ecosystem remain significant challenges. 

   • Safety: politicians and regulators expect the industry to maintain an “120%” safety threshold. That can’t be. See point #5 for my thoughts on this. 

   • Regulation: in most countries, regulation is either absent or incomplete. Where regulation exists, it is not standardised globally, hindering companies' ability to scale. Additionally, regulations often restrict projects to small-scale pilots, making it difficult to learn from large deployments and achieve commercial scale.

   • Ecosystem: There’s a lack of "native" OEMs offering market-ready autonomous vehicles. In addition, Insurance companies, due to insufficient data, quote high prices, creating further obstacles. 

3. Driver shortages and better use of time push the industry forward. 

   Focus has shifted from solving congestion to addressing driver shortages and enhancing time efficiency.

   • Driver shortages: the average age of an HGV driver in Europe is 47, with a significant proportion over 55 years old. Young people aren’t entering the profession, leading to a projected shortage of 500,000 drivers in Europe by 2026. This shortage is a significant driver for autonomous transformation in both trucking and buses. 

   • Better use of time: while not solving congestion, autonomy can make time wasted in traffic more bearable. People could watch TV shows, apply makeup, and handle emails or TikTok while riding.

4. Can Tesla change the AV industry (elephant in the room #1)

   Tesla wasn’t mentioned by the speakers (thank you, Lukas), despite deploying the largest autonomous fleet globally, with nearly 1 million vehicles using Full Self-Driving (FSD). Tesla remains an outsider due to (legitimate) criticism over tech performance and the company’s disregard for safety issues, which translates into accidents and deaths. 

   Elon Musk plans to reveal Tesla’s robotaxi in October (delayed from August). Public perception surveys suggest it will be adopted by 10% of the public on day one. If successful, Tesla's robotaxi could disrupt the industry's "safety first" and "slow and steady" approach, potentially leading to widespread autonomous vehicle use.

5. We / Society / Politicians must accept that robots might kill people. There is no way around it. We, as a society and policymakers, must accept that autonomous vehicles might cause fatalities. If we can’t accept this, we should abandon AV development. Yes, we have to make sure vehicles are safe, safer than humans, but also remember that there is never 100% (safety).

6. China’s AV growing dominance (elephant in the room #2)

   China was also barely mentioned. China has a much more permissive and risk-tolerant attitude toward AVs, as the country seeks to lead the global tech industry, and is willing to "take chances" on the way. This contrasts sharply with the West’s "safety first" attitude. 

   GDPR and spying concerns will limit Chinese AV presence in the West, but how long will this last? And what about the rest of the world? Wider adoption in China could shift the industry's foundational assumptions.

7. Regulation drives pilot locations. This seems obvious, but is the main reason why California and Germany see more AV pilots than New York and the UK. Regulation is the primary factor influencing companies' choice of pilot locations. 

8. What was the name of your last Uber driver? (elephant in the room #3)

   We already have a virtually autonomous solution. Ride-hailing uses similar supply & demand algorithms used for servicing riders in autonomous fleets; has a driver supply adjustment mechanism; and riders can work, play or rest (but not sex or alcohol) while riding, already being able to make a better use of their time.
   So what would be the efficiencies created by robotaxis then? Price, to some extent, but mostly allowing people to avoid human interaction. We must acknowledge that we are already more than halfway to autonomy in terms of daily transport network operations.

Autonomous vehicles still have their faults, but with the likes of Tesla and Chinese players, the “traditional” AV industry is at a disadvantage. We need regulators to allow larger scale deployments, and support the (safe) commercial implementation of buses, shuttles and cars in the transportation network space.