Hacker News new | ask | show | jobs
by llbowers 2515 days ago
I’m sure this is a dumb question to anyone with knowledge in this field but is there any reason to not use all three together?
7 comments

Most of the non-Tesla systems do. Waymo uses lidar, radar, and cameras. The cruise vehicles I see around have lidar and radar as well, and I just assume everyone has cameras because they're cheap and easy to stick somewhere.

To be specific about waymo (to be clear I work at Google, but don't actually have any special info on this), look at the photo in [0]. The cone thing on top is a lidar, but also has cameras in the larger part under the cone. The spinny thing on the front is also probably a lidar. The fin that looks like an extra mirror on the back, and the two on the front have radar. There's also probably a forward facing radar mounted on the nose somewhere near the grille.

[0]: https://waymo.com/

So self driving cars are basically going to be really expensive for the first while as the sensors take time to reduce in price plus the computer in the back and less distance from the battery.

Sounds like a reasonable trade off. No one needs to own these cars just rent them on demand. Plus some wealthy people in the early adoption curve.

It would make sense to use all three.

Lidar used for long range. Vision used for things like colour recognition e.g. is a traffic light green/red or is an ambulance sirens on. Radar used for reversing etc where Lidar given its location might not be able to see that close.

I suspect that optical will also supplement LIDAR in cases where very precise angular resolution is needed, such as human gesture and posture recognition (which is necessary if only because sometimes humans direct traffic, but also for things like profiling pedestrians to anticipate which is likely to jump into the road without looking.) Being able to detect which way a human head is facing will at the very least be necessary, and while you might be able to read faces with LIDAR from a distance, my gut says that optical will give you better data for that.
Price. I worked briefly with teams building self driving cars in the past. Their budget for sensors far exceeded the cost of the car itself.
Of course, they were presumably using a mostly-stock car with custom niche sensor products, so that comparison would be a bit more favorable in production.
Latest projected price for the lidar unit is $500 to $1000 e.g. https://www.wired.com/story/lidar-cheap-make-self-driving-re...

No-one is expecting to use experimental $75000 Velodyne units in production.

Currently, LIDAR is very expensive, certainly too expensive to build into every Tesla being manufactured. So Musk would not be able to sell a "full FSD capability" option on his cars if he acknowledged LIDAR is useful/necessary to autonomous driving.

The number one link if you search "Tesla" on HN is "All Tesla Cars Being Produced Now Have Full Self-Driving Hardware." It's been an extraordinarily effective marketing gimmick.

LIDAR prices have declined and are going way low-budget now.
There is a $300 vacuum cleaner robot by Xaomi that uses a LIDAR. The part can be found for $70 on AliExpress.

Obviously not the same specs as Waymo's LIDAR but at least, it proves that it doesn't have to be that expensive.

According to a recent interview with pony.ai (https://www.youtube.com/watch?v=0VcpZnIg3M0) the cost to retrofit a car with all necessary sensors is $75k.

A Lidar is a significant part of that.

If $40k Tesla can do as well as $40k car + $75k of sensors (including lidar) it's economical game over. Tesla wins by a wide margin.

The $75k will drop in time, but the battle will likely happen before the price of lidars drop significantly enough.

If self driving becomes a reality robotaxis will retail for several hundred thousand dollars.
They will certainly have an economic value of multiple hundred thousand dollars, but that applies to all manufacturers. So if a manufacturer is able to produce non-LIDAR self-driving cars and sells those cars to consumers for $100,000 less than the competition, you can bet that they’ll still capture the robotaxi rental value that is there, through an app store-like agreement. Leaving the money on the table would obviously not happen, unless it was intended to drive the competition out of business.

There would probably be room for both these models (direct sales that capture much of the self-driving value and leasing), but regardless there are obviously strong incentives for a 5- or 6-figure reduction in costs.

I have no knowledge in the field, but maybe the expense becomes high?
expensive large and heavy