How can you say camera only navigation won’t work with such finality when humans manage just fine every day!
You literally have an existence proof of it working.
Using more senses allows simpler processing of the sensor data, especially when there is a requirement for high reliability, and at least until now this has demonstrated a simpler complete system.
Artificial machines rely on spare parts manufactured elsewhere, which are used by external agents to replace the worn out parts.
For an animal to have wheels, it would have to grow wheels in some part of the body, periodically, then use its limbs to detach the wheels and attach them on the axles, after removing the old wheels. This is something sufficiently complex to be extremely unlikely to appear from evolution.
Even this huge complication would be enough only for passive wheels. For active wheels there exists no suitable motor, as the rotational motors with ionic currents are suitable only for the size of a bacteria. All bigger living beings use contractile motors, which cannot be used for a rotation of unlimited angle. So active wheels would also need a different kind of motor, which can work without a solid connection between the 2 moving parts. The artificial motors of this kind use either electromagnetic forces or fluid expansion due to temperature or pressure variation. Both would be very difficult to evolve by a living being, though electric fish and bombardier beetles show some possible paths.
At every relevant level, life relies on rotating and cyclic structures coupled through continuous material exchange. The objection to wheels in animals assumes that axles and wheels must be rigid, permanently isolated parts. Biology does not work this way. Instead of discrete components joined once and preserved unchanged, living systems implement rotation through structures that are simultaneously connected, repaired, and replaced.
Cells are full of rotary and quasi-rotary machinery. Flagella are true rotating motors with stators, rotors, bearings, and torque generation via ion gradients. ATP synthase is literally a wheel-and-axle device, converting rotational motion into chemical energy and back again. The fact that these devices operate at molecular scale does not make them conceptually different from macroscopic axles; it shows that evolution favors rotation precisely where continuous repair and material flow are required.
At larger scales, joints function as constrained rotational interfaces. Hips, shoulders, knees, and vertebrae are axles embedded in living bearings, lubricated, rebuilt, and reshaped throughout life. Bone remodeling, cartilage regeneration, and synovial fluid circulation solve the very problem claimed to prohibit wheels: permanent connection combined with continuous maintenance. The difference from artificial machines is not the absence of rotation, but the absence of rigid separability.
Even limbs themselves behave as compound wheels. Gait cycles convert linear muscle contraction into rotational motion around joints, then back into translation. Tendons wrap around bones as belts around pulleys. Muscles do not rotate indefinitely, but unlimited rotation is not a requirement for a wheel; it is a requirement imposed by certain human machines. Biological wheels rotate as much as function demands, then reverse, exactly as many engineered systems do.
The claim that active wheels require exotic motors overlooks that biology already uses fields and flows. Ionic gradients are electric fields. Blood pressure, osmotic pressure, and gas expansion are fluid-based actuators. Electric fish demonstrate macroscopic bioelectric control, and insect flight shows that indirect actuation can drive cyclic motion far from the muscle itself. The distinction between electromagnetic motors and biological motors is one of implementation, not principle.
What evolution did not produce is a detachable, externally replaceable wheel, because life does not outsource maintenance. Instead, it internalizes repair, redundancy, and gradual replacement. From this perspective, an animal is not a wheeled vehicle lacking wheels; it is a dense hierarchy of axles and wheels whose boundaries are soft, whose materials are alive, and whose motion is inseparable from their growth and repair.
Life did not fail to invent wheels. It dissolved them into itself.
All the rotating parts bigger than some tens of micrometers have only a limited rotation angle, where the limits are enforced by the solid connections between the 2 mobile parts, e.g. tendons, nerves and blood vessels.
The bacterial flagella and the rotating enzymes, which are powered by ionic currents, cannot be scaled to greater sizes. Already the flagella of nucleated cells (eukaryotes) are no longer based on rotating motors, but on contractile proteins, which must be attached at both ends on the mobile parts, limiting the relative movement.
Unlimited rotation is an absolutely necessary condition for a wheel that is used in locomotion, otherwise it is no longer a wheel.
A wheel used in locomotion that would have limited rotation would be just a leg that happens to have the shape of a wheel, because like a leg it would have to be raised from the ground for the forward motion, eliminating the exact advantage in efficiency that wheeled vehicles and tracked vehicles have over legs (i.e. that backward and forward movement are simultaneous and not separated in time during a step cycle, and no energy is wasted with a vertical oscillation of the leg).
The distinction between electromagnetic motors and biological motors is definitely one of principle and not an implementation detail. The only resemblance is that both are motors.
It is true that you can claim that when analyzing both chemical reactions and the interactions between the mobile parts of an electromagnetic motor they can be eventually reduced to electromagnetic interactions. Nevertheless such an assertion is completely useless, because most things that matter to us in the surrounding world can be reduced to electromagnetic interactions. Knowing this is not helpful at all for classifying them and understanding the differences between them.
The contraction of a protein caused by a chemical transformation and the magnetic forces that appear either between electrical currents through conductors or between electrical currents and ferromagnetic materials are very different phenomena and knowing that both of them have as primary cause electromagnetic interactions is of absolutely no help for understanding how they work or for designing either kind of motors.
Electromagnetic motors that are not extremely small need ferromagnetic materials. The only ferromagnetic material that is known to be synthesized by living beings is magnetite. Magnetite crystals can be good enough for sensing the magnetic field of the Earth, but they would be a very poor material for motors.
An easier to evolve rotating biological motor would be a rotating hydraulic motor, e.g. powered by pumped blood or lymph. This could work if the wheel would become non-living after being grown, to no longer need nerves and blood vessels. However it would be very difficult for a living being to seal the space between an axis and the rotating wheel in such a way so that blood or lymph would not spill out through the interstice.
They're still the best way we know of going about the business of building a flying machine, for various reasons.
Whether it be altimeters based on radio[1] or air pressure[2], avoidance and surveillance systems that use radio waves to avoid collisions with other aircraft[3][4], airborne weather radars[5], sensors that measure angle of attack (AoA), GNSS location, attitude, etc, many aircraft (even unpowered gliders!) have some combination of special sensing systems that aren't strictly necessary to take off, fly to a destination, and land, even if some are required for what many would consider safe flight in some scenarios.
Many of these systems have redundancies built in in some form or another and many of these systems are even built into unmanned aerial systems (UASes) big and small.
[1]: https://en.wikipedia.org/wiki/Radar_altimeter
[2]: https://en.wikipedia.org/wiki/Pressure_altimeter
[3]: https://en.wikipedia.org/wiki/Traffic_collision_avoidance_sy...
[4]: https://en.wikipedia.org/wiki/Automatic_Dependent_Surveillan...
This is the keyword here, just because the other approach is harder does not mean it is impossible.
It's a decent gamble to try and do things the hard way if it is possible to be deployed on cheaper/smaller hardware (eg: no lidars, just cameras).
No. It's not a good idea. It's not a good gamble. It's stupid, and the engineers can see it's stupid. A lot of them have quit, reducing the very slim chances of it working even further.
Hint: it's not the sensor inputs that are the bottleneck!
Not my area of expertise, so I’d rather not try to predict what will work and what won’t.
Humans are fallible and we have other sensors, like hearing, or touch (through feedback on the steering wheel) that are also involved in driving.
We already have other sensors that are not vision that work with us when driving like ABS and electronic stability.
The other reason it's dumb is that adding LIDAR and proper sensor fusion makes things better and the cost of LIDAR is rapidly dropping as its installed across new fleets in CN and elsewhere.
I never understood why they would choose to fight with "one hand behind your back". More sensors = more better
But if FSD had the same rate, people would be losing it.
I don’t think it makes sense to limit yourself while you are still figuring out what really works. You should go with a maximum of sensors and once it works, you can see what can be left out.
I think it's got something to do with an innate belief to self-determination. It's fine if I make a mistake to kill myself, and it's not fine if someone else does. It's super not fine if someone dies at the hands of a rich person's technology. Outrage, lawsuits, "justice."
CCD and CMOS sensors can easily work in sub-freezing temperatures with various kinds of heating. There are 10's of millions of surveillance cameras installed outdoors in sub-freezing climates that work fine.
Cameras also have moveable IR cut filters, which is analogous to your sunglasses example.
Human eyes do have greater dynamic range in the visible light spectrum, but solid state sensors can commonly interpret light above 1000nm, and of course you can do thermal/IR imagers to provide optical sensing of wavelengths outside of what a human can see.
Sensor technology relative to the human eye isn't what is holding FSD back.
Eyesight isn’t the thing. Humans have a persistent mental model of the world, and of the physics that drive it. Our eyes only check in every now and then to keep our model up to date.
Our ears and sense of touch do a lot of work in walking and driving, too. Trying to narrow it all down to vision is silly.
I knew a guy with no arms who drove with his prosthetic hooks. Of course he can feel vibrations and things through his ass, but so could the car if they wanted. Do they use accelerometer data? (I don't know the answer to that) Do they have ABS sensors that can detect wheel lockup/speed status? Because I don't.
I believe I can drive a car to the legal standard, remotely, with a good enough camera array.
You should fix that. Go out on a rainy day and slam the brakes hard enough for it to kick in. There's an obvious vibration and knowing what it feels like might save your life.
Of course I can feel the ABS work.
I'm saying my body doesn't have wheel rotation sensors that can act in place of ABS