Post

Conversation

At Cruise I built robotaxis that completed 250k driverless rides, including the first rides ever in a major city. With the Tesla event coming up, many people have asked for my thoughts. I truly hope they get robotaxis working. Autonomy is the fastest and most effective way to eliminate car accidents, and we desperately need this. I don’t know what we will see tonight. What I know is that it takes a non-trivial amount of work to go from making a car mostly drive without interventions to safe, robust, and legally compliant robotaxi network that meshes well with local communities. So here are 15 key things to look for from new robotaxi players: 1) Getting stuck - is there a method for remote operators to relocate unoccupied vehicles that are blocking traffic or emergency vehicles? What happens if there are no available remote operators and the vehicle becomes stuck? 2) Detecting collisions - is there a high recall system that can detect collisions, including minor contact with cycles or pedestrians? Does this comply with local, state, and federal reporting requirements? 3) AI override - is there a way for remote confirmation of critical, long tail decisions (is it safe to stop on these railroad tracks, was that a collision or not, is it ok to enter this emergency scene, is that officer telling me to stop)? 4) First responders - is there a way to remotely unlock cars and provide access to first responders who need to relocate a stuck vehicle? Is there a training program in place? Is there live phone support? 5) Connectivity dropouts - do these cars have Starlink, cellularly redundancy, or some other way to remain connected? What happens if telemetry is lost for an active but unoccupied vehicle and it gets stuck? 6) Sensor cleaning - is there a way to clear blocked or dirty sensors? Are various forms of degraded performance detected and mitigated? 7) Degraded states - in the event of a computer, sensor, or software failure, can the vehicle utilize a backup system to safely pull out of traffic or otherwise reach a safe stopping location? Is there coverage for all known or plausible faults? Will this work properly on highways where pulling over is not necessarily a safe option? 8) Congestion control - will vehicle take diverse routes to avoid creating traffic jams? If 20 vehicles arrive at a busy concert venue is there any “air traffic control” to ensure they don’t get stuck in a cluster? 9) Emergency vehicle detection - do vehicles correctly pull over or yield to emergency vehicles? Can they traverse active emergency scenes when necessary or appropriate? 10) Long tail detections - do they avoid flooded areas, downed power lines, wet cement, caution tape, crossing guard hand motions, open pits or manhole covers? 11) Liability - who is at fault when a vehicle causes property damage or injury? Are there high recall data logging systems in place as needed to absolve the owner of liability when the other driver was truly at fault? 12) Regulation and permitting - will these vehicles operate in states with public reporting requirements and meet all requirements? If vehicles do not have a steering wheel or traditional controls, will they be self-certified by Tesla as FMVSS compliant? 13) Bad weather - does the system correctly degrade its performance in the event of sudden changes in weather? What happens if a trip is in progress when weather becomes severe? Will it refuse to operate if conditions are too severe? 14) Pullovers - does the system avoid pulling over in bus stops, restricted areas, or in front of private driveways? What happens someone needs the car to be moved from their driveway? 15) Local laws - does the system obey local traffic laws, and who pays the ticket if there is a violation? Tesla will undoubtedly solve all of these eventually. Have they done it yet? Who knows. I’m still very excited to see what they’ve been cooking up.
David Watson 🥑
Post your reply

So many of those have already been answered and are already handled by what is in customer hands today. Eg connectivity dropouts. Connectivity isn’t needed. The vehicles can drive without being connected as everything on the car is sufficient to navigate. It might not take the
Show more
Thanks for sharing an objective look at things. Too many people dismiss and insult Tesla’s approach because it doesn’t fit in the nice little box that they think is the “right” way to do things. Ultimately, there will be different hardware and software approaches to solving
Show more
Square profile picture
The world's top innovators, scientists, and entrepreneurs come together at XPRIZE Visioneering to solve humanity's greatest challenges. Join us virtually to help shape the future and use code SOCIAL400 to register for just $100!
0:29
Attend XPRIZE Visioneering Virtually!
From xprize.org
What about threats from bad people (both inside and outside the vehicle)? If a passenger gets caught up in a bad neighborhood/situation, or if the passenger themselves is a bad person, is there a means reducing the threat that the taxi becomes a crime scene? It may not be a
Show more
Love the list! My additions: "School zone -- when children present" speed limits. "End road work" detection for end of temporarily reduced speed. Both of those are not handled well today, and require some judgment.
My prediction is Cybercab will be slated for a 2026, maybe even 2027 release. FSD is super dope...and works VERY good on "standard" streets. Handles everything around it very well. Buuuuut...once you get outside of vanilla roads and turns... * I live in an area where they
Show more
Great experience based considerations. I know as a Regular Tesla FSD owner that all your logging and remote access concerns are solved , (Bookmarked)
Given this list, I am still amazed that people just assume that privately owned AVs will be a thing before robotaxi fleets have been firmly established. It makes so much more sense for a fleet owner to take collective responsibility for all of this - if they can solve all these
Show more
another question - how does the system interact with other humans in ambiguous situations, whether pedestrians, cyclists and other drivers, such as communicating right of way
This is a good list of issues confronting the robotaxi concept. Additional include: A. Does Tesla expect fleet operators to buy and operate? B. Does Tesla itself plan to produce and operate a fleet, incurring the capital costs itself? C. Do they expect individuals to buy and
Show more
For #9 (Emergency Vehicle Detection) as well as hearing other driver’s honks, I’ve long thought Teslas need an external microphone. Human drivers hear sirens and react long before they see the emergency vehicle.
🔥 Relive History in Motion! 100-Year-Old Photos Come Alive with #KLING AI! 👏Elon Musk-Approved AI Platform! One-click access to our official website to unlock more creative features for FREE!🚀 Where Imagination Meets Reality: Only at KLING AI! 💫
🤖 Build Powerful AI Agents with Momen 🤖 No code, no limits—automate complex tasks with our newly launched AI feature! Build your full-stack AI apps today!
Hi - FYI The team saw your list last night at the event. They also told me that some items on the list are in QA and the others are well understood and will be addressed in short order.
This list has so many analogies for many types of autonomous services. A common theme is seamless handover to “human in the loop” modes when needed.
SpaceX Falcon 9 Axiom 3 Launch Delay I was challenged to make a prediction regarding the legendary and nearly flawless Falcon 9 Block 5 rocket launch, and I accepted. I predicted that the 3 launch would encounter difficulties if it proceeded as scheduled,
Show more
Image
Image
Image
Readers added context they thought people might want to know
This astrologer ad claims the launch was delayed for unspecified "more favorable" conditions for the Falcon 9 Block 5. However, Spaceflight Now reported the delay was due to issues on Crew Dragon: spaceflightnow.com/2024/01/17/spa… The scientific community has determined Astrology has no predictive power: en.m.wikipedia.org/wiki/Astrology…
Do you find this helpful?
Rate it
Context is written by people who use X, and appears when rated helpful by others. Find out more.
At Cruise I built robotaxis that completed 250k driverless rides, including the first rides ever in a major city. With the Tesla event coming up, many people have asked for my thoughts. I truly hope they get robotaxis working. Autonomy is the fastest and most effective way to
Show more
Great list, I have no doubt that Tesla has many of these and more in their internal systems! It’s clear they’ve thought about this future for a long time.
Bro, if they've cooked as hard as I hope, they won't even be a car company anymore What does this mean? Their multiples will be closer to that of tech companies We're looking at a whole new valuation ball game Strap in
I worry about vandalism. Self Driving job losses are an emotional issue. Certain markets won’t take it well. A bicycle and can of spray paint could bring traffic to a standstill
Thanks for sharing your thoughts...still I can jump in my #Tesla Model 3 or my Model X ...and travel anywhere in the US with Supervised FSD 98% of the time.... no cruise ever has done this.
I respect your desire for innovation and rooting for those in the arena! Nice to see and thought provoking list.
I’ve never heard about a human getting stuck and needing remote operators in order to look around an maneuver away. It’s the difference between a series of instructions and a brain. Tesla are building the latter.
16) price. they combine everything they have learned and are able to sell the robotaxi almost at cost. my prediction..... the price is going to blow everyone away.
“Getting stuck” is not a concept humans worry about. A robot trained on billions of miles of human driving will not need to worry about this either. It’s the difference between solving this problem with IF THIS : THEN THAT logic vs what Tesla have built.
Automatic driving by cars are probably the most strange way of doing driving-automation. They have been struggling for 20 years now, probably for 10 more, even with AI implementations due to causality. The solution is simple, but patentable.
Can a Right Wing-coded rideshare company even survive when cities are so important to the business model? Seems like their ceiling would be much lower than a Waymo or Cruise
I don't think the solution to "fixing" car accidents is a self driving vehicle reliant on escalating issues to a remote operator. A person needs to be liable, a human being. Show me a Robotaxi Ai that can obtain a driver's license and be able to share insurance information.
You have no idea what you are talking about. Your points show that you still don't get the point of true autonomy with real intelligence. Most of your points are moot
"Connectivity dropouts" correct me if i'm wrong but the tesla fsd model runs locally on their dual redundant inference chip on the car itself. so besides, gps they probably wouldn't need any other connectivity right?
One way sensing from car is not enough , the roads and the rest of vehicles show 2 way wireless mesh network like BLE Mesh or any long range wireless network. The regulators need to upgrade from ex traffic lights to smart lights , road sign boards to smart signboard
can't understand why car neural net needs to process data on sotounding cars , there could be 2 way transmission of telemetry data of all moving vehicles on the road , the tech is already there and cheap !
Square profile picture
The world's top innovators, scientists, and entrepreneurs come together at XPRIZE Visioneering to solve humanity's greatest challenges. Join us virtually to help shape the future and use code SOCIAL400 to register for just $100!
One thing im getting from this list is that you might not be following the FSD story very closely. Seems like a thing you might want to pay attention to
Fun list! Thanks for putting it together. It seems like you should spend some time in the current publicly available version of FSD. Give it a shot and let us know how this list changes.
There’s a better way actually - check out what’s going on in Paris & London! Cities moving away from car dependency and towards active transportation, public transit and more people spaces. It not only lowers accidents, it makes living in the city a whole lot better.
I want to see Optimus prime helping Warner bros make movies and such . There’s a reason it’s being staged there. Very exciting.
1) Getting stuck - yes 2) Detecting collisions - no 3) AI override - no 4) First responders - yes 5) Connectivity dropouts - yes 6) Sensor cleaning - yes 7) Degraded states - yes 8) Congestion control - yes 9) Emergency vehicle detection - no 10) Long tail detections - yes
/elon scribbling down these points "Tesla will undoubtedly solve all of these eventually." - laughs in CyberTruck
The other thing is supporting an offline experience. If you no longer have access to the internet, what are the risks?