Video summary
日本でロボタクシーが始まらない、本当の理由。やればできる、ただやらないだけ。
Main summary
Key takeaways
Summary of the video (auto-subtitles, key arguments and analysis)
The presenter (Andy) reviews recent global robotaxi/self-driving-car developments—especially Tesla’s “Cybercab” and London’s robotaxi hailing via the Uber app—and asks why Japan still hasn’t started widespread robotaxi service. While he acknowledges that technology readiness and public anxiety matter, his main claim is that the bottleneck in Japan isn’t primarily regulation or liability rules (which he argues are already set). Instead, it is whether companies are willing to commit to operating at full Level 4 without a safety driver handling responsibility.
1) What’s happening globally (Tesla, Waymo, London/Wave)
Tesla: “Cybercab”
- Tesla held a launch event for a steered/wheel-free, pedal-free “Cybercab.”
- Andy emphasizes that the key issue isn’t simply that the car lacks physical controls.
- The real assumption Tesla makes is that driver behavior won’t meaningfully change—i.e., the system is designed so it fully performs the driving role.
Waymo
- Andy compares Tesla’s approach with Waymo.
- He frames Waymo as still retaining a “driver/driver’s seat” concept, making it less like true autonomy than Tesla’s pitch.
London (UK): Wave + Uber
- Andy describes a service where robotaxis can be hailed via the Uber app in a collaboration involving Wave (with Nissan).
- However, he stresses it’s not equivalent to fully unoccupied autonomy:
- it uses licensed drivers/supervision
- and it involves technical/legal steps that delay “driver removed” operation
- (he cites a reporting timeline like “by Christmas,” depending on the source he references)
2) Andy’s thesis: Japan is “ready,” but isn’t “doing it”
Andy lays out five questions he plans to answer about bringing Tesla-like robotaxis to Japan:
- Can a steer/wheel-free robotaxi be legally registered and licensed in Japan?
- Can Japan permit Level 4 autonomous operation with passengers and paid service?
- What are the real obstacles (which he claims are manageable and already addressed)?
- What changes are needed for operation in Japanese cities?
- Can Japanese society accept robotaxis that operate without occupants inside?
Headline conclusions (three upfront)
- The Cybercab could be difficult for Tesla to deploy quickly in Japan due to company/operating commitment.
- Robotaxis can be implemented in Japan, but rollout will likely be “Japanese style”—shaped by constraints rather than matching San Francisco’s model.
- Most importantly: Japan already has the legal/systems groundwork (registration pathways, Level 4 permission mechanisms, and business frameworks).
- The remaining gap is practical: who will take responsibility and actually run the service at scale.
3) The four “hurdles” (and why he says they’re not regulatory excuses)
Andy argues there are four main hurdles, each with an existing solution in Japan.
Hurdle A: Vehicle approval/registration for no-steering/no-pedals designs
- He claims Japanese vehicle safety standards were updated (notably around 2020).
- These updates supposedly created pathways for autonomous-driving-related equipment, including special approvals/relaxed-standard concepts and remote-operation certification.
- He points to prior demonstrations in Japan (e.g., driverless bus/bus-like trials).
Hurdle B: Level 4 “specific automated operation” permissions
- Andy argues Japan does not allow blanket citywide freedom in the same way geofencing-style approaches might.
- Instead, Level 4 is permitted only under specific, approved conditions:
- certain roads/areas
- certain speeds
- certain environments
- He contrasts this with the U.S. model, where companies may self-certify and deploy earlier, then adjust later.
- He says Japan has issued Level 4 authorizations, with 10+ locations having permission/experience—often on fixed routes with constraints (bus-like service).
Hurdle C: Turning Level 4 into a paid passenger business
- Andy says amendments to the Road Transport Act/regulatory framework allow automated transport operators to run business using technology equivalent to Level 4.
- Therefore, in his view, carrying passengers and charging fares is not the main obstacle.
Hurdle D: Responsibility/liability in accidents
- Andy claims liability rules are already defined.
- He cites a government research process and a past conclusion (which he says dates back “8 years ago”) that Level 4 responsibilities remain with operators/vehicle owners/transport operators as cooperative parties.
- He also describes an insurance flow:
- victims are compensated promptly
- insurers investigate and may recover from manufacturers if the autonomous system is at fault
- He rejects the idea that responsibility is unclear.
- Instead, he argues the real issue is that no market actor is taking on the risk and operational burden required to run robotaxis without a safety officer.
4) Why robots aren’t starting yet: the “company commitment + market acceptance” problem
Andy returns to two missing elements:
- A company willing to seriously commit to operating robotaxis in compliance with Level 4 responsibility rules—not just trials.
- Societal acceptance, given public fear of incidents and skepticism toward unmanned vehicles.
He references deployments that still keep a safety officer in the vehicle, implying that once the safety officer is removed, accidents and responsibility become operationally unavoidable realities.
5) The role of Toyota (and why Andy thinks Toyota hasn’t led)
Andy references an earlier claim that taxi demand and fleet dominance haven’t translated into robotaxi leadership.
- He acknowledges Toyota’s autonomy-oriented projects (e.g., E-Palette / demonstrations).
- But he suggests Toyota may have avoided taking full operational responsibility in the robotaxi-service sense.
- His critique:
- Toyota may sell mobility technology/hardware
- but robotaxi “operation” requires owning responsibility for running autonomous services at scale
- which he argues Toyota hasn’t fully taken on
- He contrasts Toyota with foreign players who actually operate robotaxi services (e.g., Waymo/others), implying Japan’s delay is tied to the lack of a “serious operator.”
6) Practical/technical realism: London and “complex traffic” aren’t proof it’s impossible
To counter the argument that Japan’s traffic is uniquely hard, Andy argues:
- London is chaotic (cyclists, pedestrians, narrow streets), yet autonomy trials still proceed.
- “Uniqueness” is not a decisive barrier; autonomy needs practice and iteration.
- Drivers/robotaxis must learn unusual cases (e.g., how to behave around ambulances and fire trucks).
- He suggests AI must handle real-world edge cases under traffic/legal constraints, and performance improves through repeated exposure—not avoidance.
7) Political/cultural angle: influential people and fear of misattributed anger
Andy mentions a politician and science-fiction author (Takahiro Yasuno) who has commented on autonomous driving.
- He frames public reaction as possibly misdirected:
- if an empty robotaxi hits someone, “who do you blame?”
- He implies that as robotaxis become real, acceptance may depend on where people believe agency and blame should lie.
Bottom line
- Andy’s overall message: Japan’s legal/regulatory framework is already in place for Level 4 robotaxi-style operation (registration pathways, permissions for specific automated operation, business rules, and liability/insurance structure).
- What’s missing is:
- (1) a company willing to operate fully, and
- (2) the market’s willingness to accept it—especially when no human safety driver is physically responsible in the car.
- Therefore, robotaxis should expand in Japan, but with a rollout shaped by Japan’s permission constraints and cautious social adoption rather than a sudden San Francisco-like scaling.
Presenters / contributors mentioned (by name in the subtitles)
- Andy (presenter; creator of “Andy’s Mobility Signals” newsletter)
- Elon Musk / Tesla
- Waymo
- Uber
- Nissan
- Wave
- Nippon Kotsu
- SoftBank
- Boldley
- Wride
- Continental
- Toyota
- NTT Mobility
- Takahiro Yasuno
- “Team Mirai”