Optical tracking and laser-induced mortality of insects during flight | Hacker News Reader

Optical tracking and laser-induced mortality of insects during flight

nature.com

162 pointselsewhen120 comments

"the visible wavelengths required significantly lower laser exposure than near infrared wavelengths to disable subjects, though near infrared sources remain attractive given their cost and retina safety"

The paper repeatedly claims near infrared laser light to be "retina safe" (quotes theirs) which is an oversimplification. Infrared light of wavelength around 1 µm in sufficient powers is very dangerous for the retina, especially compared to visible light where the blink reflex can save damage by lasers up to a point. Telecom wavelengths around 1.5 µm happen to be safer because such light gets absorbed in the cornea before being focused onto the retina, but at sufficient powers or longer wavelengths (e.g. 3 µm) where the absorption depth in the cornea gets very short there can instead be damage to that part of the eye. It's always the combination of the wavelength, which determines the absorptivity in the different parts of the eye, and incident energy (pulse duration × power), that determines the safety; there are no innately safe wavelengths close to that part of the spectrum.

Maybe the best would be to mix a weak visible laser with a more powerful near-infrared laser, so that you make people blink, and your more powerful laser is less harmful? You could even trigger the visible "warning" laser 5-20ms before you fire the killer laser.

Some powerful laboratory IR lasers come with a built-in visible tracer laser. Mainly, it's useful for aligning all of the optics. With the IR laser enabled, I'm still wearing my safety goggles.

Blink reflex time is quite a lot longer than 25ms.

Figure 5 shows J/cm2 vs mortality rate. Max mortality appear around 10 J/cm2

I found a paper that shows human eye damage occuring 50% of time at 200 J/cm2, at least at blue wavelength https://pubmed.ncbi.nlm.nih.gov/26359809/

10 J/cm2 may be an adequate safety margin especially in near IR ranges where absorbtion occurs less in vulnerable inner eye structures.

There will definitely need to be some fail-safes to prevent beam from irridating one spot for more than necessary. Either beam movement or pulse length limits, or both.

Unrelated, I'm not sure why they need cameras. You already have a moving laser, couldn't you just scan the volume with it looking for insect wing beat reflections?

I guess you'd also need to look at possibilities that the laser light with the "safe" energy density of 10 J/cm² can't get focused by accident, e.g. by glasses or windows and whatnot, giving a higher and therefore potentially eye-damaging energy density.

> Unrelated, I'm not sure why they need cameras. Machine vision stuff expects a pixel map as input. You'd have to implement some quick raster scanning step to build that, but I guess in principle it would be possible.

Yeah, laser safe with optics is a real concern. Again IR might be useful here, for some frequencies only exotic materials are transparent like quartz, salt, transparent aluminum.

Using the lasers you could measure insect wing beat frequency to avoid killing non target insects. I believe I read that in an older paper about insect laser fences, maybe these authors ignored it because it's patented

The laser lab in my university had a big warning sign on the wall that said "Warning: Do not look into laser with remaining eye"

There are no safe wavelengths at all. Any frequency light at sufficient power levels will damage some part of the eye. Mechanisms may vary. Coherent light tends to be more dangerous but just looking at the sun will do damage and that's definitely not coherent light.

For every part of the spectrum there are safe limits.

Seconded. I'm working with an IR laser rangefinder for work that can blind people out to several km with the right optics in front.

Well if the laser is used for pest control in agriculture, then the solution could be as simple as a fence, making sure that no humans are around.

How did you come by this knowledge?

I work with these lasers every day. This seems to be a decent write-up: [1].

[1] https://www.rp-photonics.com/eye_safe_lasers.html

Ah I see, thanks!

It's basic biophysics.

You didn't really answer the question.

Sure he did. Read biophysics textbooks I guess is a more direct answer?

Sometimes hacker news people finds it shocking that you can receive an education from places other than blog posts on Medium and Substack.

You mean taking a 7 week intensive code camp doesn't immediately grant me enlightenment?

Anyone else chuckle at the term "laser-induced mortality"?

Laser-induced rapid scheduled disassembly.

Ah, as in missile defense systems. [1]

[1] https://en.wikipedia.org/wiki/Boeing_YAL-1

I wonder how much discussion it took to settle on that phrasing. Or maybe there is precedent?

I had a big one at reading one phrase, "lithic braking"

Lithobraking, I think? Cf aerobeaking :)

Laser-induced unscheduled disassembly would be how the mosquitoes describe it.

I wonder if it is just heat/denaturing of mosquito polypeptides, or if electrical signals in their brian fry from too much instantaneous power.

I wonder how the SDI guys used to refer to targets getting taken out by lasers. Oh wait, they didn't have to worry about it as it never worked!

Defense Secretary Espers gave a speech a few weeks ago where he said China and Russia had directed energy weapons on satellites. This was largely ignored with all the other shit going on.

Satellites are easy. They're just sitting ducks. ICBM re-entry vehicles are much tougher.

And more likely to have an ablative heat shield on the side pointing at you. Adding a little heat to a thing already being heated a lot by the atmosphere isn't as effective as blinding a spy telescope aimed at you.

Good to hear America has espers, though. As shown in Akira, espers beat directed energy weapons on satellites.

They might have just been having fun with the title.

Not really. It sounds like Newspeak.

That or humor.

It means killing them using lasers.

You call it what it is and people do not like it.

Maybe my opinion that the title is using "Newspeak" was not that wrong after all.

I love reading interesting verbose paper titles. I’ve also thought it might be fun if they added subtitles:

> Optical tracking and laser-induced mortality of insects during flight

> Or, zapping mozzies WITH LASERZ!

Same folks from 10 years ago - https://www.youtube.com/watch?v=BKm8FolQ7jw

Had to look them up.

https://en.wikipedia.org/wiki/Intellectual_Ventures

Excellent reputation they have: "The company has been described as the country's largest and most notorious patent trolling company, the ultimate patent troll, and the most hated company in tech."

I can't think of many legitimate reasons a company like IV would want 1100 subsidiaries/affiliates/whatever [1]. That's kind of a lot, no?

[1] https://www.techdirt.com/articles/20100217/1853298215.shtml

It’s one red flag we learn about in our mandatory anti money-laundering training every year.

And the reason they didn't open-source it and save a billion people right then, is they were gonna patent it all up and commercialize it.

But I still can't buy one. What the hell, capitalism?

Probably something to do with the liability of autonomous targeting and firing of high power laser pulses outside.

So release it in the public domain and let someone in Africa make the choice between a small chance of minor vision damage vs crippling and possibly fatal malaria

This is still going to require some reasonably high power laser diodes, precision galvanometers, optics, etc. Probably $2k each to cover a very small area. They just aren't going to be that practical. Genetically modified mosquitos are the way to actually solve the problem, but that has its own risks.

A speaker cone is a precision galvanometer. If one of these, manufactured, had to be sold for more than $50, retail, they are hardly even trying.

A speaker is a precision open-loop voice coil that precisely translates input current to force that displaces a membrane. A laser galvanometer is a precision closed-loop voice coil that translates input voltage to a specific position/displacement. The mechanical components to ensure repeatability and the electrical components that 'close the loop' are what generally make them (much) more expensive than speakers.

Picking off a mosquito will require positional accuracy of say 1mm. If you want a range of 10 meters you're going to need positional accuracy of ~.1 milliradian. Most galvos have a rotational range of .3-.5 radians, so you're not going to need extreme positional accuracy relative to what's on the market. A controller, pair of quality 16 bit dacs, galvo hardware and decent power supply should be adequate. I have some of the cheap $300 ebay 2-axis galvos and those are not sufficient. You're going to need a step up and are easily getting into the thousand dollar range for the galvos alone.

Why not use the same system that laser printers use with rotating mirrors and precisely timed laser pulses. I’m not sure how accurate you could get it over 10m but it would be massively cheaper and could use parts and electronics that are already produced at scale. Cheaper means you could take more of a shot gun approach to make up for any loss of accuracy.

That would definitely work if you just wanted to shoot any bug out of the sky. Probably have to use a more powerful laser to dump enough energy into the wings to damage them.

The system in the article uses reflections from a continuous illumination by the laser to identify the insect. Basically just translates to an audio signature that can be used to to see if it’s a mosquito and even if it’s a male or female.

It should be a lot more tricky with a very short illumination by scanning laser.

What about adding more scanning lasers to get a better measurement of the location? That could also be your power increase by firing multiple lasers at the same spot. Might be safer too since each individual laser would be lower power and the dangerous spot would be localized to where the beams cross.

You are asking too much of the galvanometer, and not enough of the system it is part of. Do your own muscles satisfy your requirements? Muscles no different can put a basketball in a hoop from half-court, blind.

The return signal from progessively-focusing beam can provide the closed-loop control needed.

This actually isn't asking much of the galvanometer. Commercial systems are generally 20-24 bits and have resolution in the 10-20 microradians range...100-1000 times more precise than we need.

But ultimately the control system has nothing more than input voltage to the galvanometers to steer them, and depending on the source of inaccuracy they may or may not be able to overcome it.

When the goal is a gadget you can afford to deploy, system design is where you need to get your accuracy.

Fast, close-to-linear response over a short range can be cheap; the overall linearity you get from your $1000 gadgets is unnecessary. Fixed calibration targets could help compensate for broad nonlinearity and also drift from, e.g., temperature variation. Varying focus as you home on the target with a spiral path forgives a lot of initial inaccuracy. The system could refine its response curves with each kill, to home in faster; calibration targets might not be needed if unfocused illumination is forgiving enough.

It doesn't need a 100% kill rate in the first 10 minutes. Indeed, in a usefully big volume it gets plenty of attempts on each target. So, there is plenty of time for the system to tune itself to its own hardware.

Great points. I was coming from the mindset of DIYing an open source solution rather than building a marketable product. I've been building a portable 50W CO2 burning system for marking outdoor items and all of the above has been from my experience with the tradeoffs of components you can buy off the shelf. If you were to design the system from the ground up there is a lot of fat you can trim and leverage the overall system design to compensate for shortcomings in individual components.

To me the largest engineering challenge is actually the identification step, where you'd need to get sufficient return to positively identify the critter. Would be interesting to see how they pull this all together.

You seem very knowledgeable about engineering these types of solutions. What is your background?

The reason I ask is because after seeing the OP video (which is remarkable) I am left thinking that in terms of getting a cheap MVP to the market maybe a ballistic system of some sort would be preferable.

Take the same approach of using a high speed camera but using a pneumatic launcher shoot a handful of grains of sand at high speed. There are obvious downsides to this approach but the risk to human health is lower. Thoughts?

I took a degree in engineering decades ago, but I have had this in the back my mind even longer. I haven't done anything with it, but have finally begun dinking with microcontrollers.

I would expect a sandblast to be worse to get in your eye than a flash of light. So, you would still need good large-body sensors. You would need to lead the target, another complication, and it would be hard to know how close you had got. The ones that miss would keep going and come down somewhere, although not fast. The range would be very limited because of how fast the grains would slow down -- drag goes by the area, but kinetic energy by the mass, hence volume, so smaller projectiles lose it very fast.

Not the person you asked but you'd probably be better served by looking at ways to stick with directed energy but find ways to converge it at that point so that the energy density +/- a few feet from the target is safe. Like a lower power version of this https://youtu.be/WAI7Lu4UFi4?t=825 basically long distance frying of ants with a magnifying glass.

Sand grains suffer from a large surface area relative to their mass, so their speed would drop off very quickly. I have one of those bug-a-salt guns and they kinda suck at anything more than 10 feet (also bugs seem to be pretty resilient to that kind of damage). Obviously you could make one that is more precise but I feel like there would be a lot of new variables to deal with.

Some kind of beam steering ultrasonic setup might work as well (although probably expensive) If you can get 10-15 transducers to pop off a precisely timed 5W burst of say 60kHz sound such that they constructively interfere at the critter, you might be able to get them to disassemble in air without any moving parts on your rig.

That seems like a winning strategy: an array of ultrasonic transducers could both detect and range the bug, and then you just play back a loud impulse spike with the same phase delays as you heard. No moving parts, no misses, no calculations. If the spike is short enough, you physically can't damage anybody's ears.

Maybe you only deafen them, so they die of old age without mating.

I once priced out an array of half-inch-sized ultrasonic transducers. Seems like they were astonishingly cheap... like under $.50 each? Plus $.50 microcontrollers and a drive transistor for the spike. Maybe an FPGA running parallel convolutions?

"Genetically modified mosquitos are the way"

So you think a massive invasive change in the core biology of a very large animal population is clearly superior to a local solution?

Ok, then please explain why.

Personally I would much rather leave the wildlife genetics as it is and rather protect homes directly. (btw. mosquitos are annoying even if they do not carry malaria)

It wasn't a value statement. I don't think it's the right way to go about it, but it would almost certainly be the most effective. These laser boxes aren't really a 'local' solution. They would reduce, not elimiate, mosquitos in the vicinity of one entry into a home. You would need a dozen to fully protect a single house.

When you describe one solution as "the way" it implies it is the only rational way, which is pretty much a clear value statement.

And since the area covered of the laser is 30 m, killing up to 100 mosquitos per second, why on earth would you need a dozen to protect one home?

I mean, it's "outside" in a very loose sense. There's an evenly-illuminated backlight to identify the suckers in the first place, which can act as a backstop for the high-power beam. It's almost as contained as a desktop laser cutter, and a lot less power. (Or at least shorter shots, anyway.) It's not like you're firing beams into the sky or across the street or something.

Solid work, but I find it quite concerning that there appears to be no discussion of potential false positive detections of other species beyond those specifically targeted. The research is still in a very early phase, and `wing-beat frequency, and/or other factors` means they are thinking about it a bit. If I deploy one of these I want to know how many friendly or benign insects I'm going to be killing.

True that, but if a system of this sort is not used, I think in reality it would not be the case that nothing would be used, but rather that very broad-spectrum insecticides get used. That's what happens now. So even a somewhat-targeted method would be a big improvement over what happens now.

They are not "thinking about it a bit", their earlier work was concerned with precisely identifying the species and gender of the mosquito.

I believe this system can very accurately ID and neutralize an exact specie of a mosquito.

yeah, from 10 years ago:

> The laser detection is so precise that it can specify the species, and even the gender, of the mosquito being targeted. “The women are bigger. They beat at a lower frequencies,” Mr. Myhrvold said. Since it is only the female mosquitoes who bite humans, for the sake of efficiency, his system would leave the males alone.

-- https://bits.blogs.nytimes.com/2010/02/12/using-lasers-to-za...

I work near a restaurant kitchen, and they spend much of the day with a flyswatter in hand getting rid of these pests. I have actually verbalized this idea to the chefs in the past, I'm glad to see it actually being done. If made cheap enough this could be a real solution to keeping bugs away without resorting to nasty chemicals or netting everywhere.

This has the potential to be a big win for not only comfort, but health and safety as well. Less bugs in kitchens = less chance for food getting infected or worse yet, having eggs laid in it.

I am told that bead curtains in doorways keep out (or keep in) flies. It seems surprising, but they are very heavily used in some places, so there might be reasons to believe it.

I wonder what the mechanism is. Do the beads need to be transparent?

I've seen this in the Mediterranean quite a lot. Particularly in Spain. Beads do not need to be transparent, and it works well.

Why are they infested with flies? If you're having to swat them you're already doing something wrong.

What's the point in verbalizing your laser flyswatter idea to chefs, when even most engineers wouldn't be able to build one?

"Dude, it would so awesome if they like, built a laser torrent to shoot down these things" "Yeah, that would be sweet" "Yeah"

Because it's a neat sci-fi idea and you're talking to a colleague...

If memory serves me well, 6 or 7 years ago a former microsoft employee made such a device. It ran on solar power and had just enough power to melt the wings of mosquitoes. I wanted to buy it ever since I saw the prototype, but I dont think it was ever released.

We tried to schedule some meetings with them a few years ago, but after some back and forth they were difficult to deal with. In any case, prior to the meeting I had done some preliminary research, there is other interesting research in the area

Here's a video of tracking using, if I recall, a fairly cheap camera - https://youtu.be/kuaMcVf501Y?t=4

And here's one of the associated papers - http://cdcl.umd.edu/papers/rsi2012.pdf

This paper in particular is interesting because, at the time I was working with an optics expert, and one of the major limitations at the consumer level was what this new paper aims to addresses (power output and safety). One observation we made was that, for consumers there are not too many laser 'things' in the home that are not fully enclosed. Basically, the FDA takes laser safety extremely seriously. People often think only about looking directly into a laser, but backscatter can also be dangerous.

These are the same people. Nathan Myhrvold is who you're thinking of -- he was the former CTO of Microsoft, and among other things, now runs an outfit called Intellectual Ventures, who did both that work several years ago and is responsible for this paper (I wasn't sure at first, but there are references to Intellectual Ventures Labs further down into the paper). IV is either a research institution of a patent troll or both, depending on your perspective, but either way, they don't generally commercialize the ideas they come up with, they just license them.

Intellectual Ventures did the initial work on this project, and while commercialization languished for quite some time, an org somewhere in the Microsoft alum web ("Global Good", supported by the Bill and Melinda Gates Foundation) is moving forward with commodification of the technology [1]. The patent doesn't expire until 2032 [2], but I've always had the thought that an open source version would not easily be constrained by patent laws (especially if the open devices were deployed in developing countries, where needed most) (IANAL).

[1] https://photonicsentry.com/

[2] https://patents.google.com/patent/US8705017

So this invention which could remove a significant portion of human suffering is being squatted on? And because it is patented no one else can remove this suffering either?

What a nightmare. The patent system is enormously harmful to the world and should be done away with.

They're called Intellectual Vultures in some circles...

Depends on whether you think people would work nights and weekends to get an idea out the door if someone with more money can immediately copy and profit.

History says no, but maybe there is a different system.

How many sucessful patents are from freestanding inventors nowadays actually?

The main in-action of patent I have seen is patents being hoarded in big corporations. If the engineers work nights and weekends at those it is their own fault.

They are rewarded with salaries that have never before existed outside of law/business/rent seeking (kings/land owners/etc.), the main power brokers of wealth in most of human history.

Not necessarily, or yet, but it would certainly fit IV's MO if it were.

Maybe this could be considered prior art: a satirical Dutch commercial from 2007 (or earlier) about the "Star Wars Mosquito Defence System" https://youtu.be/TGkPMZxWPpA

> IV is either a research institution or a patent troll or both, depending on your perspective, but either way, they don't generally commercialize the ideas they come up with, they just license them.

That’s not exactly accurate. IV creates spin-out companies that commercialize particular inventions created by IV inventors. Photonic Sentry is the commercialization vehicle for this mosquito zapper that was invented by IV inventors. Other examples are Kymeta, Terrapower and Echodyne.

IV does not typically commercialize inventions (patents) that it purchases — but the patent purchasing piece of the business is distinct from the invent and commercialize piece.

The 4 videos at the end of the article at https://www.nature.com/articles/s41598-020-71824-y#Sec18 spark joy - these beasts are somehow attracted to my scent/blood so I allways get the bites while others are spared.

Your contribution to humanity is appreciated by the rest of us being sparred the bastards going after you. Fight well my friend!

Is your blood type o-negative? Any time mosquitos are out i get bit more than anyone else by a margin.

It's the blood group O that draws mosquitos, not a positive or negative rhesus factor: https://pubmed.ncbi.nlm.nih.gov/15311477/

I'm O-positive and am always the center of their attention. It's amazing, sometimes people don't even believe me when I tell them how often I'm being bit (because they're getting nothing).

Interesting. I just heard something about that in passing and it made sense. My father in law is type AB (and a lifelong smoker) and mosquitos completely ignore him. I'll have ten bites and they don't even get in his vicinity..

same here. I wonder whether it is related that the O-negative is universal donor - i.e. it has lowest chance of causing immune reaction by not having all those antigens/factors, and thus i'd speculate lowest chance of "indigestion" in mosquito :)

Wrt. the original post - if we put "mortality inducing" laser on a drone so it would collect the "fried" insects sparrow style then we'd have a nice protein rich dinner after half-an-hour of the drone flying around. Turning the tables so to speak.

Careful, mosquitos will evolve to be reflective if we deploy enough of these :-D

"Transparent enough to not be detected by these systems" is probably more likely. ;)

Only if we wound and not kill

No. The very killing applies selective pressure.

They don't seem to know that there is a specific optical wavelength that dissociates chitin via a mechanism not dependent on heating. It seems like much shorter or less focused exposures should suffice at such a wavelength.

In particular, a mosquito whose wings are shredded does not need to be (also) killed to be eliminated as a threat to public health.

Sounds very interesting. Can you provide a link?

I haven't looked it up in decades, but it is what I was thinking of long before IV built and patented their thing.

It would have to be in fairly deep ultraviolet or insects could not fly in daytime.

It seems really interesting at first glance, but having worked with high-speed cameras, filming a mosquito at 30m at 2000fps will only be practical in a bright sunny day. Some clouds or at night and you will need kW's of light just for tracking the mosquito. So the emphasis on power budget for the laser seems pointless to me.

This needs to exist.

Mosquitoes apparently have accounted for half of all human deaths ever.

https://www.nature.com/news/2002/021001/full/news021001-6.ht...

This was presented at a TED talk 10 years ago (2010)

https://www.ted.com/talks/nathan_myhrvold_could_this_laser_z...

What's different now?

Salt-induced mortality of flying insects: https://www.bugasalt.com/

Reminds me of the Africanized bee zapper from David Brin's Earth:

http://www.technovelgy.com/ct/content.asp?Bnum=1790

Looks like this particular paper is focused on the laser part. I wonder how well the tracking part of it can work—if the bee zapper is practical.

Machine vision system that detects insects and targets them with a retina safe laser, however:

> For both the coarse and fine tracking systems, subjects are identified by the size of their silhouettes generated from near infrared LED back-illumination or reflection.

Very impressive system but it currently requires a uniform backlight.

How is this different from this project from 2011? https://magazine.columbia.edu/article/what-bugs-mosquitoes

"Szabolcs Marka, an associate professor of physics, is developing a novel way to protect people from the vectors: a virtual mosquito net with infrared light vibrating at wavelengths that irritate the insects’ nervous systems."

Awesome stuff. I'd like to see entirely non-chemical pest control for agriculture.

God bless scientists. We might not have flying cars yet but this is almost as good.

I used to admire Nathan Myhrvold for his work on the Modernist Cuisine and this, until someone cut him down in a HN comment that just because he plays with lasers and mosquito's in a lab and makes cookbooks it doesn't excuse his patent trolling and damage to the world.

And to be honest the Modernist Cuisine isn't as good as the hype I've since decided after actually reading it.

Anywho if this ever leaves a lab I might re-evaluate my opinion on him.

There were more than one occasion when I hunt for the mosquito sneaked in the bedroom that I had thoughts of building a laser gun to zap the sucker.

In theory you could use a lower powered beam to cause the insect to turn around and stay out of certain areas. Or use microwaves instead.

God here.

So species 1 (in this case, mosquitos) is annoying species 2 (in this case homosapiens). I'm rather disappointed to know that species 2 has taken it upon themselves to maim species 1 by burning them. I may need to do something about this.

This is straight out of Cory Doctorow's book 'Pirate Cinema', can't wait until hats with these become commodities and I can repurpose them to knock out CCD security cameras!

Kudos to Intellectual Ventures Laboratory, Nathan Mhyrvold, Matthew D. Keller, Bryan J. Norton, David J. Farrar, Phil Rutschman, Maclen Marvit & Arty Makagon! I look forward to deploying this tech to deal with fruit flies soonish. I don’t even need to do a back of the envelope calculation to have confidence this is soon to be a viable first world niche product. I think it’s likely viable at a massive scale (many millions of units) too.

Infrared lasers are NOT retina-safe.

are not _necessarily_ retina-safe.

Interesting subject, but the experiment is a rube-goldberg mess--a cartesian product of frequencies, intensities, durations, and a rather complex targeting system.

Not looking forward to the same kind of rig being attached to high-power lasers, and pointed at civilians.

Not that I think there aren't governments and other organizations willing to do that, but it has been proven time and again that chemical explosives (e.g. gunpowder) is just hard to beat for killing efficiency. In other words, anyone willing to do that, would probably be willing to just shoot them with a bullet or drop a bomb on them, and that would always be cheaper.

explosives don't really jive with the social credit system's API

> Not looking forward to the same kind of rig being attached to high-power lasers, and pointed at civilians.

Already happening: https://en.m.wikipedia.org/wiki/Active_Denial_System

Reply on news.ycombinator.com