The Disrupters: Some LiDARs could cause blindness, warns Aeye

Aeye’s Luis Dussan makes the case against 900-nanometre LiDAR to Louis Bedigian.

For all the talk of road safety, you don’t hear many people discuss eye safety when it comes to autonomous vehicles. LiDAR, a key ingredient in many autonomous vehicle plans currently underway, is really just a spinning laser sitting atop each automobile. While they might be deemed a harmless threat, history tells another story that has caused at least one start-up to proceed with caution.

Luis Dussan, founder and CEO of AEye, is in charge of a company that is developing a vision hardware, software and algorithmic solution for AVs. Like many start-ups in the auto industry, AEye is using LiDAR but Dussan, an aerospace and defence veteran, was adamant about not using lasers in the 900-nanometre wavelength, which poses a number of potential hazards. “You don’t know you’re looking at it and the next thing you know you’ve injured your retina and it’s permanent damage for the rest of your life,” Dussan warned. He explained that 900-nanometre lasers, which are considered to be eye safe by the International Electrotechnical Commission (IEC), caused a lot of injuries within the US military. Even professionals who knew the risks – these lasers have proven to be dangerous when viewed from just 10 centimetres away – incurred permanent eye damage. The military ultimately found that lasers in the 1550-nanometre wavelength were much safer. They became known as “retina-safe.” Thus, AEye chose to use 1550 nanometre LiDAR instead.

Dussan is surprised by the ongoing use of the former, however. He noted that with binoculars (or any other magnifying device), 900-nanometre lasers are dangerous at much greater distances. This is especially problematic if the laser stops spinning but continues to beam in one particular direction, which strengthens its potency. Most LiDAR are being designed to shut down when they aren’t spinning but, if they malfunction, they could be very dangerous to the naked eye.

Said Dussan: “The IEC standards say that’s probably rare, you’re not likely to be harmed, so it’s okay and that might work for one or two lasers on the road on a given day but, when there are millions of lasers on the road, that is absolutely not going to work. How are you going to prevent your kid from going up to a car with binoculars? You can’t. You also can’t prevent them from putting their eyes on that system while it’s spinning.”

No risk beyond testing?

Dussan does not expect the 900-nanometre wavelength to go beyond the testing phase for AVs. He said that while the military may have been able to get away with it from a liability standpoint, commercial industries, such as automotive, will not be that lucky. “Laser eye safety is extremely important and the companies out there that are doing 900-nanometre lasers, they do not know the dangers associated with that laser,” he said. “We don’t use it and we don’t think it’s going to stand the test of time, or the court of public opinion.”

Some companies may be unwilling to switch but Dussan wonders if that could change through regulation. He said: “900-nanometre lasers are also not safe with respect to the infrastructure. They are right in the middle of the band of all the night vision/low-light cameras that are being used for ADAS. It’s a dumb, silly way to design the system. Let’s put out vehicles that are going to blind the cameras that are being used in other vehicles! With 1550, you don’t have that problem because cameras can’t see 1550. There’s no conflict there.”

Yet, that’s not the only risk. Dussan said that security cameras – the ones used to help law enforcement – are also in conflict with the 900-nanometre wavelength. “The government is probably going to step in – we know, we’ve talked to them – and could probably regulate 900-nanometre based LiDAR just on that alone,” he said.

Long-term exposure

Only the short-term risks are known for any laser, including those that are likely to be deployed with self-driving cars. The risks might sound trivial to those who have never come face-to-face with an AV, or those who expect to stand more than 10 centimetres away from the spinning puck sitting on the roof. There’s another risk to consider, one without any clear answers: long-term exposure.

“Nobody has ever deployed a million lasers on a highway,” said Dussan. “That really doesn’t exist, so long-term exposure to your eyes is an unknown thing. Is there some long-term exposure that could potentially hurt your eyes? I wouldn’t know an answer to that. I don’t see how that particular wavelength is going to do anything to your eye that we can see. It’s certainly no more powerful than blue light you get from the sky or anything like that. It’s one band in the entire spectrum of bands that you’re always looking at.”


2 comments

  1. Avatar Mark 3rd July 2019 @ 8:22 pm

    And what about other creatures in nature that are exposed and have retinas that are compromised at different wavelengths than we are?

    Computer driven cars are a disaster in the making, trillions and trillions in misallocated capital to develop both they and the 5G that they will require to function as intended. And combine potential health risks of LiDAR with with that of 5G and you have a massive threat to health on a scale that nobody has ever imagined.

  2. Avatar Sorce 10th July 2019 @ 2:48 pm

    The problem is that even with 1,500 nm lasers to be honest we’d need more research to know the possible damage they could do in certain very rare cases such as with replacement artificial eye lenses, for certain types of animals with very thin lenses… I, myself, had a short encounter with a Waymo on the street in San Francisco and while looking at the spinning thingy on the top I got an unpleasant sensation in my eye. Now nothing happened to my vision, everything works as it should and maybe it was all placebo or whatever but some serious research is warranted.
    If LiDAR turns out to be dangerous in the long run, or has adverse effects on the environment, or otherwise bad news, this would turn a huge amount of investment money to [rubbish]. Multiple companies would have to start again from scratch or go bankrupt. Let’s hope people will do the right thing and research this the right way, human safety first, money second.

Leave a comment

Your email address will not be published. Required fields are marked *