Open Acoustic Devices

@OpenAcoustics

Creators of the #AudioMoth. Follow us for updates on the device and our work in conservation technology.

Joined February 2017

1.4K Following

4K Followers

3.2K Posts

OpenAcoustics retweeted

Modern Explorer

@ModernExplorer5

26 days ago

A peer-reviewed acoustic localization framework just hit sub-5-meter precision using low-cost GPS recorders — and our research found that number intersects directly with consumer GPS watch accuracy in a way that has real field implications. Freeland-Haynes et al. published the framework May 9, 2026 in Communications Biology (doi: 10.1038/s42003-026-09949-5). The pipeline is fully automated and open-source, built on synchronized GPS-AudioMoth recorders. In loudspeaker field tests, 99% of broadcast calls were localized to within 5 meters. The system resolves multiple simultaneous sound sources, rejects error without manual review, and was demonstrated at scale — an array of over 60 recorders surveying birds across a forest site. Code is on GitHub. Audio test data is archived on Zenodo. The team at University of Pittsburgh, led by Louis Freeland-Haynes with co-authors Sam Lapp, Chapin Czarnecki, Lauren Chronister, Joe Begley, Alex Rogers, Peter Prince, Andy Hill, and Justin Kitzes, addressed a persistent bottleneck: acoustic localization has historically been constrained by hardware cost and effort-intensive manual analysis. This pipeline removes both barriers simultaneously. Our research found the cross-domain connection worth stating plainly: 5-meter acoustic localization accuracy is directly comparable to the positional precision of consumer-grade GPS sport watches. That is not a coincidence to file away — it is a structural match. Wildlife survey grids built on this pipeline could be co-registered with recreational trail GPS tracks, meaning any hiker, trail runner, or backcountry traveler carrying a GPS watch is already operating at the spatial resolution required to serve as a reference node in a passive acoustic monitoring network. Documented evidence suggests this makes opportunistic citizen-science localization networks technically feasible right now, using hardware that already exists in the field. The strange part: the limiting factor for large-scale passive monitoring of unidentified vocalizations in remote terrain was never the recorders. It was the analysis pipeline. That barrier is now gone. An investigator could deploy a grid of inexpensive AudioMoth units in a documented anomalous-activity zone and receive precise spatial origin data on every vocalization — with no one on site. What do you know about this?

ModernExplorer5's tweet photo. A peer-reviewed acoustic localization framework just hit sub-5-meter precision using low-cost GPS recorders — and our research found that number intersects directly with consumer GPS watch accuracy in a way that has real field implications.

Freeland-Haynes et al. published the framework May 9, 2026 in Communications Biology (doi: 10.1038/s42003-026-09949-5). The pipeline is fully automated and open-source, built on synchronized GPS-AudioMoth recorders. In loudspeaker field tests, 99% of broadcast calls were localized to within 5 meters. The system resolves multiple simultaneous sound sources, rejects error without manual review, and was demonstrated at scale — an array of over 60 recorders surveying birds across a forest site. Code is on GitHub. Audio test data is archived on Zenodo.

The team at University of Pittsburgh, led by Louis Freeland-Haynes with co-authors Sam Lapp, Chapin Czarnecki, Lauren Chronister, Joe Begley, Alex Rogers, Peter Prince, Andy Hill, and Justin Kitzes, addressed a persistent bottleneck: acoustic localization has historically been constrained by hardware cost and effort-intensive manual analysis. This pipeline removes both barriers simultaneously.

Our research found the cross-domain connection worth stating plainly: 5-meter acoustic localization accuracy is directly comparable to the positional precision of consumer-grade GPS sport watches. That is not a coincidence to file away — it is a structural match. Wildlife survey grids built on this pipeline could be co-registered with recreational trail GPS tracks, meaning any hiker, trail runner, or backcountry traveler carrying a GPS watch is already operating at the spatial resolution required to serve as a reference node in a passive acoustic monitoring network. Documented evidence suggests this makes opportunistic citizen-science localization networks technically feasible right now, using hardware that already exists in the field.

The strange part: the limiting factor for large-scale passive monitoring of unidentified vocalizations in remote terrain was never the recorders. It was the analysis pipeline. That barrier is now gone. An investigator could deploy a grid of inexpensive AudioMoth units in a documented anomalous-activity zone and receive precise spatial origin data on every vocalization — with no one on site.

What do you know about this?

194

OpenAcoustics retweeted

GroupGets

@groupgetscom

28 days ago

https://t.co/98vTOfr2DM 😸 Wildlife will not need to look both ways when crossing across one of Asia's largest wildlife elevated corridors standing at average of 7 meters/23 feet high. The National Highway Authority of India (NHAI) collaborated with the Wildlife Institute of India to make an underpass on the Delhi–Dehradun Economic Corridor using 29 @openacoustics Audiomoths. 😸 Wildlife will not need to look both ways when crossing across one of Asia's largest wildlife elevated corridors standing at average of 7 meters high. The National Highway Authority of India (NHAI) collaborated with the Wildlife Institute of India to make an underpass on the Delhi–Dehradun Economic Corridor using over 40 @openacoustics Audiomoths. The 20-km-long stretch between Ganeshpur and Asharodi features a total of 10.97km of animal underpass designed specifically to facilitate unobstructed animal movement. The area of the study covered flat riverbed, hilly terrain and mixed stretches of Sal Forest in the Shivalik Himalayan range, through a 40-day monitoring programme using 150 high-tech camera traps and 29 AudioMoth acoustic recorders. The study documented a total of 111,234 images of humans, domestic animals, and wildlife. Among these, 40,444 were attributed to 18 unique wild species utilising the underpass, including various carnivores, herbivores, ungulates, pheasants and primates. The golden jackal was the most frequently captured, followed closely by nilgai, sambar and spotted deer.

groupgetscom's tweet photo. https://t.co/98vTOfr2DM 😸 Wildlife will not need to look both ways when crossing across one of Asia's largest wildlife elevated corridors standing at average of 7 meters/23 feet high.

The National Highway Authority of India (NHAI) collaborated with the Wildlife Institute of India to make an underpass on the Delhi–Dehradun Economic Corridor using 29 @openacoustics Audiomoths.

😸 Wildlife will not need to look both ways when crossing across one of Asia's largest wildlife elevated corridors standing at average of 7 meters high.

The National Highway Authority of India (NHAI) collaborated with the Wildlife Institute of India to make an underpass on the Delhi–Dehradun Economic Corridor using over 40 @openacoustics Audiomoths.

The 20-km-long stretch between Ganeshpur and Asharodi features a total of 10.97km of animal underpass designed specifically to facilitate unobstructed animal movement. The area of the study covered flat riverbed, hilly terrain and mixed stretches of Sal Forest in the Shivalik Himalayan range, through a 40-day monitoring programme using 150 high-tech camera traps and 29 AudioMoth acoustic recorders.

The study documented a total of 111,234 images of humans, domestic animals, and wildlife. Among these, 40,444 were attributed to 18 unique wild species utilising the underpass, including various carnivores, herbivores, ungulates, pheasants and primates. The golden jackal was the most frequently captured, followed closely by nilgai, sambar and spotted deer.

182

Open Acoustic Devices

@OpenAcoustics

Who to follow

Last Seen Users on Sotwe

Trends for you

Most Popular Users