Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Pot hole avoidance?

This site may earn commission on affiliate links.
I now know AP1 doesn't have pot hole avoidance!

I didn't expect to see such a large pot hole on the Gardiner Expressway Eastbound and neither did AP1.

Wondering if AP2 will have the capability?

What happens when there's road debris such as items falling from trucks? Has anyone experienced these scenarios?
 
Seems to me, that simply imaging a road surface, over time, should reveal the presence of a pot-hole -- as distinguishable from debris, which is normally transitory. Tesla should federate its data sharing with other forward-camera data collectors (BMW? Mercedes?) so a) cars can avoid; and b) local authorities do triage to tackle the problems in timely fashion.
 
...Tesla should federate its data sharing with other forward-camera data collectors...

LIDAR can see 3-d so it knows if the image it senses is flat, shallow or deep. It can even sense the thickness of the paint for lanes.

Tesla without LIDAR cannot.

However, Tesla does have a neuro network where data is constantly fed to the headquarter which will help 10/2016 hardware.

It would recognize that there's a specific image on the road at a specific location that drivers keep reacting by slowing down, or steer around it.

Its Artificial Intelligence program would then learn to mimic drivers' reactions and download its conclusion to the rest of Tesla fleet that each time it encounters that location, it would expect there's an image that it should slow down or steer around it.
 
Last edited:
  • Like
Reactions: Missile Toad
LIDAR can see 3-d so it knows if the image it senses is flat, shallow or deep. It can even sense the thickness of the paint for lanes.

Tesla without LIDAR cannot.

However, Tesla does have a neuro network where data is constantly fed to the headquarter which will help 10/2016 hardware.

It would recognize that there's a specific image on the road at a specific location that drivers keep reacting by slowing down, or steer around it.

Its Artificial Intelligence program would then learn to mimic drivers' reactions and download its conclusion to the rest of Tesla fleet that each time it encounters that location, it would expect there's an image that it should slow down or steer around it.
It seems human eyes without lidar can detect potholes just fine (for the most part)

I would argue that there are numerous benefits to using a camera to detect potholes instead of lasers.

Here's a paper that agrees with me Pothole Detection System Using a Black-box Camera. - PubMed - NCBI

Recent automatic detection systems, such as those based on vibrations or laser scanning, are insufficient to detect potholes correctly and inexpensively owing to the unstable detection of vibration-based methods and high costs of laser scanning-based methods. Thus, in this paper, we introduce a new pothole-detection system using a commercial black-box camera. The proposed system detects potholes over a wide area and at low cost....

Wondering if AP2 will have the capability?
No way to know unless you're on Tesla's AP2 team... but I can tell you for certain that it's a software issue.
 
I suspect pothole is another specialty that requires further research and proven records.

Google & Jaguar are working on it with additional pothole sensors (including suspension sensors) but some one might have to sacrifice their tires, wheels, axles first so the cloud system can record how bad a pothole is.


Google patents pothole-detecting technology - Car Keys

googlepotholebody3.jpg
 
I suspect pothole is another specialty that requires further research and proven records.

Google & Jaguar are working on it with additional pothole sensors (including suspension sensors) but some one might have to sacrifice their tires, wheels, axles first so the cloud system can record how bad a pothole is.


Google patents pothole-detecting technology - Car Keys

googlepotholebody3.jpg
I'm not sure I appreciate Jaguar's approach as it only works if someone actually hits the pothole. Sometimes it's unavoidable but it's far better to detect and report without having to hit it. Google's patent is virtually the same thing as it's composed of GPS and a vertical accelerometer.

Ford has a different approach: pothole mitigation:


Mercedes uses an optical camera to detect the road surface ahead.

 
Pothole avoidance would be easier to tackle than flying debris. An example today I was driving on the Burlington Skyway which is known for severe winds, when an 18 wheeler covered with snow and ice on the roof, came flying off in sheets. There was no way to avoid it being in the passing lane howevever how about further behind? Wondering how AP would react to that if at all.
 
  • Informative
Reactions: Jaff