We are incredibly thankful to WILDLABS and Arm for selecting the MothBox for the 2024 WILDLABS Awards. 

The MothBox is an automated light trap that attracts and photographs moths and other nocturnal insects. A raspberry pi (mini-computer) controls a super high-resolution camera and lights, so that the MothBox can be deployed and programmed to collect data at a pre-defined schedule. A computer vision model then processes the images and automatically identifies the insects captured by the trap.

Moths are one of the most diverse orders of insects and a great indicator of the overall diversity in an area because moth larva (caterpillars) are often highly specialized, feeding exclusively on single host-plant species, or living in narrow niches like rock faces, moss, or decaying wood. As adults they are important pollinators, and their populations support birds and other small predators. However, despite the diversity and ecological importance of moths, they are seldom included in conservation planning due to the difficulty in including them in monitoring systems. Conventional monitoring of moths involves moth-sheeting or setting out bucket traps, which require many hours of tedious work by scientists with specialized taxonomic knowledge. This is the problem the MothBox and other automated camera traps for insects seek to address.  

These are many groups working on automated camera trapping of insects (as is evident from the WILDLABS group), and we intend to continue and strengthen our collaboration with these groups to make insect monitoring more accessible and scalable. What sets the MothBox apart is our focus on a low-cost, low-weight, jungle-proof solution that can easily be deployed en masse by non-scientists in remote tropical locations.

Currently, parts for one complete MothBox cost approximately $450, though we are working on ways to decrease this cost further. The MothBox is fully open source with in-depth documentation on Github: https://github.com/Digital-Naturalism-Laboratories/Mothbox

My partner @hikinghack  has been documenting the MothBox project on WILDLABS for a while, see our past posts/updates here:

• Introducing the MothBox: https://wildlabs.net/discussion/cheap-automated-mothbox
• Update 2: https://wildlabs.net/discussion/update-2-cheap-automated-mothbox
• Update 3: https://wildlabs.net/discussion/update-3-cheap-automated-mothbox
• Update 4: https://wildlabs.net/discussion/mothbox-mothbeam-update-4
• Also, checkout the recording of @hikinghack k presenting the MothBox during a wildlabs event back in November: https://wildlabs.net/event/automated-moth-monitoring-you

With the WILDLABS Awards funding this project can really take off. Some of the things we plan on doing in the coming year: 

• Adding capacity for a solar panel or additional batteries to extend deployment time.
• Deploy MothBoxes at sites around Panama, doing a national moth inventory and collecting big data for training AI identification models.
• Hire AI and taxonomic experts to create, validate, and refine computer vision models.
• Work with the Panamanian reforestation NGO Pro Eco Azuero to quantify biodiversity uplift associated with native-tree planting and natural regeneration of tropical forests.

We would also like to thank all of the people and organizations that helped get us here: Earthshot Labs for the initial funding to get V1 off the ground, Michigan State University which helped develop a drone-deployable version of the MothBox for the Rainforest XPRIZE competition in Singapore last year, Experiment.com which funded development of V3 of the MothBox, ETH Zurich which provided funds for construction of 11 MothBoxes in late 2023, and of course our partner @mothyash  for his valuable insight on moth behaviour and raspberry pi programming.