alextousss

Reminds me that comma.ai has its own line. While their setup is quite expensive they do run production for a smartphone-level of difficulty in-house. They detail it incredibly well at commacon [0].

[0]: https://www.youtube.com/watch?v=1iiBSN224w4

[1]: https://www.youtube.com/watch?v=LE-QlIWMHxM

Had a chance to see a live demo last month. Looks great, can do a lot more things than an SO-101 and the teleop via the app if both fun and useful. Would definitely buy if I had the money.

That will end up either dangerous or unreleased. Their system sends a 40W laser, which is what you use to cut plastics. There’s no way to make this safe as even a diffuse secondary reflection out of the field of view could blind someone.

Exactly! Plus we’ll get a nice beautiful sonar echo when they’re on walls/ceiling

Long term, around $300/unit

Unfortunately it won’t. an eye that collimates light is much more fragile to laser than a mosquito.

Everything works in simulation, and the detection works in the real world! We’re working on miniaturizing the electronics to embed it in a real drone.

Mosquitoes use different cues (olfactory, CO2 and infrared emissions/heat) depending on the distance to target.

We could do something about it. Any flying insect is only a change of parameters for us.

What do you mean by fidelity? We’re still unsure about the tails in terms of detection probability/false detection rate

It might be, but it’s a crowded market where politics and connections are worth more than technical expertise. So, we’re okay to stay out of it

Yeah, it’s one of our goal to work for government agencies at some point to implement city-wide mosquito control. 10 of our drones could cover a square kilometer, so we’d be a lot more efficient working at the city level rather than at the individual household level.

Not our highest priority concern haha

Their velocity is much lower than the one of the drone, so it wouldn’t make much sense to increase efficiency

Yeah, you’re spot on! The original name was « mosquitorgnole »

Just an insane obsession with ultrasound! (and some control theory classes)

Glad to hear we could be of help! Some of our tech could be used for defense, but traditional defense companies and ukrainian startups already do low-cost shahed interceptors.

Yeah, what they do is very cool. Not sure how far they are in the development but the videos are super cool.

My roommate and I are still working on Tornyol, our mosquito killing drone! It uses ultrasonic sonar to detect mosquitoes, and missile control theory to ram into mosquitoes and grind them in its propellers.

Our target platform is a 40 grams tinywhoop so it’s safe to fly everywhere and makes almost no noise :). A Roomba for mosquitoes!

The main plus compared to traditional systems is that a drone can cover an enormous surface in a short time compared to static systems or man-portable insecticide spraying. Our goal is to be competitive with ITNs against Malaria.

Some links :

https://hackaday.com/2025/03/25/supercon-2024-killing-mosqui...

https://manifund.org/projects/build-anti-mosqu

Light being coherent and light being dangerous are two completely orthogonal concepts. A laser is dangerous because of the very tight beam, but any household mirror keeps that beam focused. Plus, a mosquito needs 100millijoules to fry (from the intellectual ventures study) which is a lot more than what a retina can handle. The short duration actually makes it even worse, because the eye lids can’t close soon enough to avoid the retina burning (they take 100ms to close). That’s the reason why lasers are classed by power.

A retina being easier to burn than a mosquito is the fundamental reason why we haven’t seen that tech deployed. LiDAR detection with a 2D goniometer + high power laser has been available commercially for a while. There just never was someone to create something so deceptive as to sell a 40watts device as possibly safe.

I may haven't fully understood your answer, but a typical household mirror could reflect 90% of the laser power in a single coherent direction. Any sufficiently polished metal tool would have dangerous specular direction. I'm not sure of the math for a diffuse reflection, but the laser classification is here for a reason.

A human eye being transparent up to the fragile retina, yes, a laser would penetrate the eye and be concentrated in an extremely small spot on the retina. That's exactly the reason why we have safety around lasers, and why everything above 5mw is strictly for enclosed use. 40 watts shot at random in the void is definitely dangerous by all measures.

I’m not here to sell my own startup, but there are other ways to kill mosquitoes than a 40 watts laser that could make someone blind from a secondary reflection (out of the mmWave radar detection range)

They do in a YouTube video on their channel [0]. It’s 40 (!) watts. They also show it shooting indiscriminately at any small object, with an example where they feed it foam chips.

[0]: https://youtu.be/Ta0f0oB4I-Q?si=WqlhTmVHVszXSjtU

Tornyol | AI Control Research Intern | Embedded Systems Intern | ONSITE (Paris) | full-time | 2K€/month | https://tornyol.com

Tornyol is building anti-mosquito drones to fight against malaria. We use smartphone microphones to detect mosquitoes, than a policy learned with reinforcement learning to guide our drones towards mosquitoes. We're in the middle of raising a pre-seed, and are looking for two interns to join us for six months starting in September.

Here are the roles and requirements :

- AI Control Research Intern: code, test and deploy reinforcement learning-based drone control policies and simulation environments. We're looking for a background in either reinforcement learning or classical control.

- Embedded Systems Engineer: Design the physical platform. Responsibilities include everything from CAD to designing PCBs or writing embedded signal processing code.

Perks:

- Small but extremely talented team. The two founders are very technical and will help and guide you in all your missions.

- Lots of responsibilities right from the start. It's a startup.

- Pay is great

If you think you could be a great fit, please email a resume + a few lines on previous personal projects you've built at the email in our website's footer.

I’m interested in how to secure the firmware from reverse-engineering in a product. Is there somewhere an introduction to the techniques used by SpaceX there?

Incredible article, learned quite a lot. To me, a very good supplementary reading would be Structures by J. E. Gordon [1]. Helped me grasp a lot of the mechanical design notions necessary for that sort of work.

[0]: https://archive.org/details/StructuresOrWhyThingsDontFallDow...