Mothbox 4.0

https://digital-naturalism-laboratories.github.io/Mothbox/

by bookofjoe • 2 years ago

277 42 2 years ago

digital-naturalism-laboratories.github.io

morninglight• 2 years ago

This could be a great teaching tool, BUT at about $375 per Mothbox, it's going to be a hard sell for most schools.

hansihe• 2 years ago

$375 per box doesn't seem bad to me at all when you probably only need a couple per school?

hikinghack10• 2 years ago

Yeah, i feel like currently they are at about the price of camera traps 10 years ago. There is very little mass-manufacturability to them right now (it's all open source and made from off-the-shelf parts) but later if we can find more funding, we are going to make a design more for manufacturing which should hopefully drive the costs down even more! :)

yapyap• 2 years ago

I don’t think it’ll be a hard sell for schools at that price, school budgets tend to be fairly elastic for new things (and sometimes for wholly unnecessary things, for the price of less than 1/4th of new MacBook for the principal (for example) they could have this.

Edit: also if schools were to be interested in this (which they should be, it’s very neat) they could group together and buy in bulk which would also greatly reduce the price

nerdponx• 2 years ago

Are there any "citizen science" initiatives where I can buy a device like this and upload my insect counts to some database that's useful for researchers?

I have been interested in setting up some things like home weather monitoring, ADS-B, streaming webcams for wildlife, etc. anyway so this would be a fun item to add to that.

gibspaulding• 2 years ago

Cornell is doing something along those lines for birds using audio recordings. It’s just a smartphone app though - https://merlin.allaboutbirds.org/sound-id/

beardyw• 2 years ago

I was using it today, and any day when there are birds around. It correctly identifies birds you can barely hear. It's become as useful as binoculars to me. A great free app.

dekhn2• 2 years ago

they also release the model. I've recorded hours of audio outside my house and run it through the model on my desktop with GPU.

buescher• 2 years ago

I have a pi running birdnet-pi on my porch. https://www.birdweather.com/birdnetpi

adolph2• 2 years ago

Here is the related hardware widget. $200 with a $59/year service for alerts etc.

https://haikubox.com/

aeturnum• 2 years ago

I think the air gradient folks do some collaboration and it's possible you could get your sensors linked into a bigger net by contacting them: https://www.airgradient.com/research/

Edit: Also MIT did this recently: https://news.mit.edu/2023/low-cost-device-can-measure-air-po...

Modified30193• 2 years ago

On a related note, I wish there was some sort of software system where I could ID weeds and insects with photos and locations at various levels of confidence (“possibly”, “probably”, keyed out, genetic testing) at different taxonomic levels and both self host my results while also automatically contributing to a larger project.

There are various projects out there (like plantnet) but I don’t want to burn massive personal effort curating in a system that isn’t my own first and foremost, due to inevitable enshittification. At the same time, I want others to benefit from what I do, in particular local growers and naturalists. Things like PlantNet also tend to be “majority vote” on ID, meaning a whole lot is often close, but wrong. For example there is a regional plant specific to my area called Willamette Navarretia. Those that don’t realize this will easily confuse it with very similar looking species found most elsewhere in the western US, and last I checked it wasn’t in PlantNet.

https://www.oregon.gov/oda/programs/PlantConservation/SiteAs...

https://www.inaturalist.org/taxa/165663-Navarretia-willamett...

doodlebugging• 2 years ago

Have you tried iNaturalist? [0]

Insects, animals, plants, etc get posted from all over the world for ID.

[0] https://www.inaturalist.org/

I post lots of photos of things I find locally and experts step in and help ID the subjects when I don't know what I'm seeing.

I also have a couple of AudioMoths for recording local sounds including birds, insects, etc. Very high quality units at a reasonable price.

I have seen this Mothbox posted a couple of times and almost bought one since I know I have tons of moths attracted to my native plants out here, but the price is a sticking point right now. I think something like this combined with an AudioMoth and some trail cams would be a good local wildlife monitor setup.

xipho• 2 years ago

If you want to manage your own data take a look at TaxonWorks. Multi-entry and bifurcating key editors, images, much more. It, and its companions are all open source, see TaxonPages for example.

throwup238• 2 years ago

due to inevitable enshittification

PlantNet is led by an alphabet soup consortium of French research institutes like IRD, CIRAD, INRIA, and INRA so I don’t think enshitification is inevitable - though I don’t know why it’s not open source. It’s funded by grants and donations with no incentive to enshitify it for more revenue.

r2oaa2f83jfpa3• 2 years ago

Not counts, but you can upload the images to https://inaturalist.org. Most of the recent conservation research papers use iNaturalist data.

smallerfish2• 2 years ago

BioClip (referenced in their writeup) is good to know about.

I run a small scale wildlife monitoring program with usually 3 cameras (depending on how many are being repaired for battery terminal corrosion). Each month I skim through and categorize ~200 videos that we've captured. Doesn't take too long, but it would be nice to automate at some point.

Modified30193• 2 years ago

Do you use any kind of protectant, like dielectric grease or Deoxit? I find applying to both the equipment and battery contacts saves a lot of headaches on my work phone, ATV, trailer lights connectors and anything else powered I use in the field.

smallerfish2• 2 years ago

No, but great idea. Once I get the current batch back from the repair shop I'll start doing that.

hikinghack10• 2 years ago

We use some dielectric grease when we have to make some outdoor connections and it works great here in the jungle, and it also has the bonus of keeping the leaf cutter ants from slicing up your electronics too!

Choco31415• 2 years ago

BioCLIP won the CVPR best student paper award: https://cvpr.thecvf.com/Conferences/2024/News/Awards

Having talked with Sam, they're working on BioCLIPv2. So expect even better results sometime soon.

hikinghack10• 2 years ago

Hi! I'm andy! I'm the guy who spent the past year making the Mothbox! I don't know how this got on ycombinator, but cool! We are working here together Let me know if you have any questions!

Some quick basics: It's an entirely open source system made from off-the-shelf parts made so you can build one yourself! We built it to help a set of community reforestation initiatives here in Panama! (

Existing automated insect monitors we checked out wouldn't work for us because: 1) They were REALLY BIG (like huge heavy suitcase size+) 2) They were really EXPENSIVE (like 7000-15000$ !) 3) There wasn't any info to build robust ones that could withstand months in the rainforest!

So we build one ourselves! We started with absolutely no budget, and the costs that people have noticed ($375) are really the cheapest you can find these parts (Pi5 + big battery + 64MP camera + really bright lights), and for instance are on par with the $200 just for the 16mp logitech webcam employed in many existing insect monitors!

We got about 20 of these out in the wild here in Panama already, and are literally building more in the jungle as I'm typing this!

steve_adams_862• 2 years ago

Hey Andy, this looks awesome. Nice work!

I think I'd like to get one to track insect diversity and populations in areas where I work on riparian restoration projects. My hope is that when projects are completed, in the following years it'll be evident in the insect data the Mothbox collects. Do you think that's a valid use case? Maybe I'd need more than one to get enough data?

hikinghack10• 2 years ago

Totally! We are using it for farmland and riparian restoration projects here in Panama. Everything is still very nascent, so we don't really have a good idea what amount of mothboxes would constitute "enough data," but the interesting thing about insects vs like mammal camera traps is that like even if two mothboxes are just 200 m apart, you can really see a lot of insect variation!

bookofjoe• 2 years ago

Here is how you got on ycombinator:

1. I happened on your website in Clive Thompson's newsletter, among the websites in his final "sudden death" round:

https://buttondown.com/clivethompson/archive/linkfest-25-the...

2. I thought HN peeps would find it of interest so I submitted it

hikinghack10• 2 years ago

Awesome! Thanks for sharing it around!!

chrisweekly• 2 years ago

Thanks for sharing -- and even more, for the Clive Thompson link. :)

mason_mpls• 2 years ago

thank you for your service

hikinghack10• 2 years ago

Thanks! Always happy to serve the commons of science!

sinuhe69• 2 years ago

I skimmed through the user manual and find the design is really cool. It is well thought for the field use and obviously the optimized result of many iterations. Although I'm no naturalist, I find inspiration in your design. Kudos to the good work!

hikinghack10• 2 years ago

Really appreciate the kind words!

adr1an• 2 years ago

I wonder if any predators (i.e. birds) would eventually learn about this lure and its schedule, and starts waiting around it for its preys...

Mothbox 4.0

Related Stories