Hacker News Reader

Comment by yoodenvranx | Hacker News Reader

I know almost nothing about Lidars and I was unable to find an answer to my question:

What happens if you have a few dozen Lidar-equipped cars in the same spot? Is there something like crosstalk, scatter or noise between the cars which result in Lidar artifacts?

View on HN

According to [1] mutual interference would be unlikely because the beam and sensor are tightly-focused, and there's a very short time window in which a spurious beam would affect the measurement. [2] mentions that radar units modulate the waveform to identify the one it sent out, I imagine the same thing could be done for lidar if interference becomes a problem.

[1] http://ideas.4brad.com/tricking-lidars-and-robocars [2] http://www.aiqus.com/forum/questions/22373/using-multiple-li...

I believe they can send out a unique fingerprint of pulses so that UnitA is not confused by UnitB signals. Much in the same way that you could have three cars in a row, but each one only responds to the remote control for it.

The lidar is only going to be listening for a pulse less than .005 of the time since light goes very fast and you only get so many points per second. I've had a small fleet of robots running on lidar in the same room and crosstalk just hasn't been a concern.

Reflected sunlight caused us a number of problems but I'm going to assume that any lidar used on a car is rated for outdoor use. I suppose those might still give bad readings if you're driving directly into the sunset or something, though.

Not quite as bad as relying on visual sensing while driving into the sunset as a white truck crosses in front of you...

For Velodynes you get some spurious points that look like noise cuased by reflections or rain/snow. Any mature system already has some kind of filter so it's tractable. The rate is on the orders of 10's to 100's per second which is low compared to the million/second of the sensor.

I think usually you're OK - the scary stuff with lidar is what happens when it snows or it's foggy. Or even worse, a malicious person with a beam that they point into your sensor.

Well, to be fair, a malicious person could just as easily shine a beam into the optical sensors on my face.

If someone really wants to cause an accident, it's kind of hard to stop them, regardless of whether you're in a self-driving car or not.

The good news is humans also can't see well in snow or fog, computers should be able to at least do as well plus better know their limits and not drive beyond their abilities.

Highly unlikely unless such interference is very carefully, deliberately planned.

If one did want to think adversarial one could explore a system receiving Lidar, say from a car behind, figuring out its transmission and doing your own illumination to confuse or jam the other car system. But the chances for this state arising naturally are astronomically small. My 2c.

I am not 100% sure about that. At the moment it is basically impossible that there is interference from another LIDAR due to the fact that there are so few. but if every car, motorcycle, drone and whatnot is equipped with several LIDARS then interference might become real. Or it could at least cause the noise floor to be raised to a level where sensor readings are affected.

But I am not saying that this will become a huge issue, because it would affect individual measurements, and the overall scan should be fine, just a bit noisier. And noise is something they have to deal with today as well due to different reflectance properties, electromagnetic interference, flying voxels, etc.

Agree. Raising the noise floor is possible (but I suspect you need a lot of devices to do that perceptibly), but then it is usually relatively easy to adjust for, or at least detect the fact that you are in a higher noise environment.

There are commercially available systems for jamming speed measuring lidar used by police.[0] Although I am unsure if this would disrupt a line scanning lidar

[0]https://en.wikipedia.org/wiki/Laser_jammer

seems like you could make a material that would absorb most of the light, rather than rebounding it? Lidar uses NIR, or UV lasers, so if you could engineer a material that had good absorption properties maybe you could prevent the beams from rebounding and confuse it that way?

You don't need to engineer it. In our lab tests we accidentially realized that one of our desktop computers was "invisible" to the Velodyne, because it had a black case.

Every material that is black could be a good candidate. I am wondering if the google cars do have a problem with cars that have a matte black finish (those PVC foils).

But you have to keep in mind that they are not invisible in the real sense, because you get a hole in your scan where you don't have points in your pointcloud. This could be detected, because the Velodyne usually has many beams pointing downward which eventually would hit the road, but would leave a hole instead if there is an NIR absorbing material.

And if you use sensor fusion, you can use a camera to check for dark colors in those areas that might help you to detect such cases

The charcoal-black material sometimes used for office chair upholstery is a very good IR absorber, and those things are not visible to a SICK LIDAR.

That's why you want to profile the road from a high LIDAR. It's not just about sensing obstacles. It's about building a height map of the environment. This was more important in the DARPA Grand Challenge, where we had to drive off-road. Potholes and rocks were to be expected. There are visual illusions for both vision and LIDAR, but few of them present the appearance of a flat road.

This is what happens sometimes, totally inadvertently.

Reply on news.ycombinator.com