You can install our site as a web app on your iOS device by utilizing the Add to Home Screen feature in Safari. Please see this thread for more details on this.
Note: This feature may not be available in some browsers.
The sensors stick out even farther than the prototype. Definitely zero blind spots on that thing.
The sensors stick out even farther than the prototype. Definitely zero blind spots on that thing.
No blind spots with occupancy network either. Superhuman.Definitely zero blind spots on that thing.
No blind spots with occupancy network either. Superhuman.
No, not necessary; just need to construct a model of the world and always be moving. As long as you could see the spot very recently it is fine. Very few possible exceptions. Obviously a child or animal could potentially run in front of the car at the last second, but would hit them due to physics even with no blind spots (and I certainly have zero confidence the existing autonomous vehicles have the ability to even brake for such things, though they might in some cases).Occupancy networks do not guarantee zero blind spots. Occupancy networks only work if the camera can see the spot. So you need sensor coverage to guarantee no blind spots.
No, not necessary, just need to construct a model of the world and always be moving. As long as you could see the spot very recently it is fine. Very few possible exceptions. Obviously a child or animal could potentially run in front of the car at the last second, but would hit them due to physics even with no blind spots.
Agreed.And you should not accept blind spots just because situations where you might hit something might be rare. It is much better to have 360 degree sensor coverage where there are no blind spots in the first place.
I won't comment on the aesthetics, but I'd say Tesla had better jump on the recognition training for these kind of double-ended People Mover vehicles. Otherwise we're going to see FSD Teslas freaking out, thinking that a van is coming at it in the wrong lane. Or at minimum, an FSD screen visualization of a van being hitched up backwards to a tow truck, much to everyone's amusement.
I think this is where we'll see divergence from L2, L3, L4 and L5. Waymo's, Cruise's, and other AVs at L4 and L5 will never be sold to the public, only available as ride-hailing / robotaxis. L2 and L3 will be on consumer vehicles for purchase/lease.Occupancy networks still require sensor data to work. So the occupancy network will not be able to update the occcupancy of an area that it has no data on. Sure, if your car is moving, the occupancy network might be able to make a prediction about occupancy of a spot that was visible a couple seconds ago that is now in the blind spot. And what if your car is not moving? That would be a very poor and unreliable way to do autonomous driving, if you are dependent on your car moving all the time, to hopefully check blind spots. And you should not accept blind spots just because situations where you might hit something might be rare. It is much better to have 360 degree sensor coverage where there are no blind spots in the first place.
I think L2 will require more and more strict driver monitoring as it gets better so taking your eyes off the road will not be an option.The L2 and L3 world will be get into the car, behind the wheel, and have it drive to a destination for us while we pay some attention, but can get distracted by phone calls, text messages, etc. and still have a safer drive than if we drove ourselves.
Weird they are getting rid of it right as we enter the glorious worldwide robotaxi utopia.Btw, Intel bought Mobileye for $15.3 billion.
Years and years later. Mobileye is IPOing for $16 billion, lol.
The sun is setting on Mobileye. Good time to get rid of it.
![]()
Mobileye IPO Would Value Self-Driving Car Company At $16 Billion
Intel stock edged up Monday, as it gets ready to spin off the self-driving technology company Mobileye, which could raise $820 million.www.investors.com
I suspect Waymo is far ahead in the losing money category.Btw, Cruise's net income for Q3 2022 was negative 500 million. This is a negative 2 billion per year run rate.
Cruise's bull-case revenue in all of 2025 is 1 billion (no mention of operating expenses), contingent on their software progress.
500 million is the cost of 10,000 Model 3s. Lol.
Good luck.
The whole point of these scammy fsd companies is to release marketing PR stunts to justify their huge losses.
Maybe they’re just setting up to be the next Tesla. Or maybe developing driverless vehicles is a high risk / high reward proposition.Btw, Cruise's net income for Q3 2022 was negative 500 million. This is a negative 2 billion per year run rate.
Cruise's bull-case revenue in all of 2025 is 1 billion (no mention of operating expenses), contingent on their software progress.
500 million is the cost of 10,000 Model 3s. Lol.
Good luck.
The whole point of these scammy fsd companies is to release marketing PR stunts to justify their huge losses.