Mosquitoes tend to be seen as an almost universal negative, at least in the lives of humans. While they serve as a food source for plenty of other animals and may even pollinate some plants, they also carry diseases like malaria and Zika, not to mention the itchy bites. Various mosquito deterrents have been invented over the years to solve some of these problems, but one of the more interesting ones is this project by [Ildaron] which attempts to build a mosquito-tracking laser.
The device uses a neural learning algorithm to identify mosquitoes flying nearby. Once a mosquito is detected, a laser is aimed at it and activated in order to “thermally neutralize” the pest. The control system as well as the neural network and machine learning are hosted on a Raspberry Pi and Jetson Nano which give it plenty of computing power. The only major downside with this specific project is that the high-powered laser can be harmful to humans as well.
Ideally, a market for devices like these would bring the price down, perhaps even through the use of something like an ASIC specifically developed for these mosquito-targeting machines. In the meantime, [Ildaron] has made this project available for replication on his GitHub page. We have also seen similar builds before which are effective against non-flying insects, so it seems like only a matter of time before there is more widespread adoption — either that or Judgement day!
26 thoughts on “Bug Eliminator Zaps With A Laser”
“so it seems like only a matter of time before there is more widespread adoption”
Sea lice zapper: https://hackaday.com/2017/04/10/submersible-robots-hunt-lice-with-lasers/
I wrote the software that powers the Stingray sealice devices. After feasibility demonstration I recruited two friends, and together we built the overall system. To solve such problems you need resolution, speed and precision in both spatial and time dimensions. For this reason I prefer to only use C and ASM and as few libs as possible.
Coming from the 8-bit and Amiga demoscene I knew something about the importance of optimization and minimalism in software engineering, and that enabled us to build a high performance system at very low cost. If anyone is interested I run a startup IN-A-VISION that specializes in robotic vision systems.
The first of these was made many years ago, but no one has ever marketed one. Most likely for liability reasons.
Whilst this is certainly an issue, I suspect that price and complexity vs simpler options are more likely the issue.
Or because Nathan Myhrvold has a patent on it?
Although, I would have expected it to have expired by now.
this is the latest (2020) I could find stemming from that project: https://www.nature.com/articles/s41598-020-71824-y
It appears that that team now has a company dedicated to this project http://photonics entry.com and are now selling the tracking part as a standalone research tool. I wonder if there are any independent demos of it, tracking 1000 insects in flight at once at a range of ~20 feet is no small thing. It would seem like the actual goal is quite achievable, affordable I don’t know.
once this gets working,it needs to be refined so that there is
spectator satisfaction,and getting the laser pulse just right so that there is a flash,with a pop sound,and then a little smoke trail,
and for £#$¥’s sake if we can give robo cars a human take permit
we can surely zap a few trillion blood sucking parisites and live with the liability
And innocently enough the first “killer robots” were put in service. If only they knew what the future would hold….
The “harmful to humans” (primarily the eyes) would be trivially solved if it is placed well above eye level, and constructed such that it cannot aim down. Then it’s mostly a matter of making sure there are no non-diffuse reflective materials within range (shiny metal, glass, etc.), and focusing the laser such that it’s harmless at long distance (airplane range).
Alternatively, mosquitos could be attracted via carbon dioxide, heat, and pheromones, into an enclosed space, but then they could be easily killed with an electrified grid, completely removing the laser from the system. That would probably work, but lose the epic “killing them with lasers” bit.
Or, you know, spray a bit of RAID in the room and put netting on the window, so that the bugs can’t get in.
But yeah, that wouldn’t have neither lasers, nor hype with machine learning …
This specific laser idiocy has been around for several years. It is originally a research project by some engineering student dude in Russia. And it can’t even detect a mosquito on a white wall (by his own admission, given the cited performance of the tracking system), much less in flight and against non-contrast background …
It periodically surfaces like a zombie out of the grave, marketed as a solution to the mosquito problem with malaria in Africa and such.
Yes, we should move onto something more practical… like Snoopy’s doghouse styled micro drones that engage individual mosquitos in air to air combat.
If you wanna develop the avionics on a more macro scaled platform first, here ya go…
I didn’t make a YT detailing… though last season I tested out two screens (intake and fan exhaust windows) where I used painters tape to wrap uniformly no-see-um netting around to the outside of the screen frame and then super glued the screen to the inside of the frame. Then the excess no-see-um mesh fabric was either trimmed or tucked back in the seam to make a nice tighter fit for the screen frame for less gaps.
Ghetto’istic, though effective and this last week I wrapped two other screens interestingly.
If you’re wondering, why didn’t he just remove the screen spline/tube and redo all together?
Well, the size spline I had on hand was smaller and wasn’t tight enough in the groove so I opted for at the time what I had available… $1 store super glue in sufficient quantity and that will tide me over until I can (or want to) afford to replace my windows.
Reminds me of the Starwars Musquito Defense System
Pew Pew Pew
There was a start-up claiming to implement this a few years back, and every once in a while some “journalist” would see it and re-surface it which I imagine gets the company a few more “pre-order” dollars…
What they did get right about in their approach is that killing the mosquito is actually not necessary. If you tell me where it is, I’d be happy to catch it myself.
The liability issue seems relatively simple.
You make 3 of them (minimum) and each is about half the necessary power.
You mount them at approximately 7′, and you have a positioning system that locks at horizontal.
Then you use a coordinated during system to pulse all at the same target simultaneously. You can do similar range lockouts to make vertical curtains at fences and walls.
If it can be feasible and work with killing Asian Tiger mosquitos, I’d buy it. They make me miss the usual brown ones that just were a minor annoyance.
To me it seems Ants would be a much more usable and tractable target
If some one can adapt this to kill varroa mite on a bee can be a nice solution
Wonder if you could do that by having the bees fly through a terrahertz laser beam, where the wavelength is close to the size of the mite, and might kill them extremely quickly while not harming the bee.
There was an attemp Some years ago by a spanisch scientist
Wot, no sharks?
Seems Paul Allen funded something along the lines… though some days I wonder if he got zapped for not being so profiteering and/or petrochemical industry dope everything minded. Anyways….
I was thinking this would be safer and possibly more effective in a farming situation on a machine where can zap weeds and bugs at the same time with less issues of safety… especially focused to target down by the easier to discriminate non-target stems.
Guessing a refined design can zap critters while at it.
I remember that post, that was so cool. Weed killers that could be used for farming which use lasers for weeding make my brain smile as well.
Please be kind and respectful to help make the comments section excellent. (Comment Policy)