Gundi tool gets conservation hardware and software talking the same language

For years, monitoring water levels in the semiarid Ol Pejeta Conservancy in Kenya was a tedious task. Team members would travel the 50 kilometers (30 miles) from the main office to the nature preserve to ensure that water levels were adequate enough for use by the animals that live there, including the last two northern white rhinos (Ceratotherium simum cottoni) on Earth.

When, in 2021, the team installed digital water meters to enable remote monitoring, they ran into a hurdle. The data output couldn’t be read by EarthRanger, a monitoring and analysis platform widely used to collate and visualize real-time data gathered from multiple tracking and monitoring devices.

It’s an issue that has plagued conservationists and organizations that use multiple devices and software to track animals, monitor the safety of rangers on patrols, and visualize enormous data sets. Popular equipment such as telemetry devices, camera traps and audio recorders are produced by multiple manufacturers from around the world. Often, it takes a lot of time, money and resources to convert data from each of these devices and integrate them into platforms where they can be stored and analyzed.

A new open-source tool aims to resolve this compatibility issue and enable conservationists, protected-area managers and researchers to deploy tools without worrying about whether the data produced will be usable. Gundi, which means “glue” in Swahili, is a result of a collaboration between EarthRanger and the nonprofit organizations Wildlife Protection Solutions and the Wildlife Conservation Society.

At Ol Pejeta Conservancy, Gundi has now helped the team access readings from the water meters. “Every day, we get to view the recordings, and this has really helped us financially by reducing costs,” Martin Mwaniki, head of technology at the conservancy, told Mongabay in a video interview.

Gundi serves as an adapter between the hardware and software. It takes data from different sources, transforms them into a common format, and then sends them to different destinations for analysis. Currently, the platform allows for 100 inbound integrations and is being used by conservation organizations and national parks in Africa, Southeast Asia and South America.

“Any conservation technology program that wants to make use of data coming from wildlife trackers, camera traps, tracking devices for rangers, along with deforestation information and wildfire alerts, can integrate with one source rather than having to do all of those integrations separately,” Jes Lefcourt, director of conservation technology at the Allen Institute for AI, where he heads the EarthRanger project, told Mongabay in a video interview. “It’s like plumbing under the street. Today, everyone is out there installing their own plumbing everywhere and there’s a lot of complexity that goes into getting the water. Gundi takes care of all that complexity.”

Currently, Gundi can be integrated with 50 types of hardware devices, including wildlife trackers, camera traps and radio telemetry devices. It also works with popular data monitoring and analysis platforms such as EarthRanger, Global Forest Watch and iNaturalist.

Lefcourt said Gundi is meant to aid those who are building new devices as well as those who use platforms that generate analyses and insights that guide conservation action. He cited a new camera trap being developed by a team at Arizona State University that also incorporates acoustic detection capabilities. Normally, the camera-trap data would need to be integrated with a platform such as Wildlife Insights, which uses machine learning to analyze the thousands of images gathered. The audio data would need to be separately integrated with platforms such as Arbimon, a popular eco-acoustic analysis software. “But now, they could do just one integration with Gundi,” Lefcourt said.

Gundi is currently being used to integrate data coming from multiple collars used to track pumas (Puma concolor) in the Patagonia region of South America. It’s also being used to track and monitor the rewilding of scimitar-horned oryx (Oryx dammah), a species whose threat category recently improved from extinct in the wild to endangered.

In the months ahead, Lefcourt said, the team plans to test and develop Gundi further to add at least one new integration every two weeks. Lefcourt said he’s also excited by the prospect of three organizations coming together to find a common solution to a conservation issue.

“We could have been working in silos, wasting a lot of time and energy in duplicating effort, and creating a more fragmented technology landscape,” he said. “But instead, we are all working towards this common goal to ensure that every dollar spent on conservation technology moves the mission further as opposed to making it redundant.”

Banner image: Gundi is being used to track and monitor the rewilding of scimitar-horned oryx (Oryx dammah), a species whose threat category recently improved from extinct in the wild to endangered. Image by liesvanrompaey via Flickr (CC BY 2.0).

Abhishyant Kidangoor is a staff writer at Mongabay. Find him on 𝕏 @AbhishyantPK.

Mongabay Data Studio: Revealing nature’s hidden stories through data (commentary)