Keeping Badgers At Bay With Tensorflow

October 15, 2023 by Danie Conradie 9 Comments

Human-animal conflict is always a contentious issue, and finding ways to prevent damage without causing harm to the animals often requires creative solutions. [James Milward] needed a humane method to stop badgers and foxes from uprooting his garden, leading him to create the Furbinator 3000, a system that combines computer vision with audio deterrents..

[James] initially tried using scent repellents (which were ignored) and blocking access to his garden (resulting in more digging), but found some success with commercial ultrasonic audio repellent devices. However, these had to be manually turned off during the day to avoid annoying activation of the PIR motion sensors by [James] and his family, and the integrated solar panels couldn’t keep up with the load.

This presented a good opportunity to try his hand at practical machine vision. He already had a substantial number of sample images from the Ring cameras in his garden, which he turned into a functional TensorFlow Lite model with about 2.5 hours of training. He linked it with event-activated RTSP streams from his Ring cameras using the ring-mqtt library. To minimize false positives on stationary objects, he incorporated a motion filter into the processing pipeline. When it identifies a fox or badger with reasonable accuracy, it generates an MQTT event.

[James] modified the ultrasonic devices so they would react to these events using an ESP8266-based WeMos D1 Mini Pro development board and added an external 5 V power supply for sustained operation. All development was performed in a Docker container which simplified deployment on a Raspberry Pi 4.

After implementing the system, [James] woke up to the satisfying sight of his garden remaining untouched overnight, a victory that even earned him some coverage by the BBC.

Thanks for the tip [Laurent]!

An image of two dogs and a bison wearing harnesses with the energy harvesting system. Text next to the animals says Dog 1 (Exp. 1), Dog 2 (Exp. 2), Dog 2 (Exp. 3), and Wisent (Exp. 4)

Kinefox Tracks Wildlife For A Lifetime

June 17, 2023 by Navarre Bartz 17 Comments

Radio trackers have become an important part of studying the movements of wildlife, but keeping one running for the life of an animal has been challenging. Researchers have now developed a way to let wildlife recharge trackers via their movements.

With trackers limited to less than 5% of an animal’s total mass to prevent limitations to the their movement, it can be especially difficult to fit trackers with an appropriately-sized battery pack to last a lifetime. Some trackers have been fitted with solar cells, but besides issues with robustness, many animals are nocturnal or live in dimly-lit spaces making this solution less than ideal. Previous experiments with kinetically-charged trackers were quite bulky.

The Kinefox wildlife tracking system uses an 18 g, Kinetron MSG32 kinetic energy harvesting mechanism to power the GPS and accelerometer. Similar to the mechanical systems found in automatic winding watches, this energy harvester uses a pendulum glued to a ferromagnetic ring which generates power as it moves around a copper coil. Power is stored in a Li-ion capacitor rated for 20,000 charge/discharge cycles to ensure better longevity than would be afforded by a Li-ion battery. Data is transmitted via Sigfox to a cloud-based database for easy access.

If you want to build one to track your own pets, the files and BOM are available on GitHub. We’ve featured other animal trackers before for cats and dogs which are probably also applicable to bison.

Saving Birds With 3D Printed Boats

May 31, 2022 by Bryan Cockfield 22 Comments

Montana, rightfully nicknamed the big sky country, is a beautiful state with abundant wide open landscapes, mountains, and wildlife. It’s a fantastic place to visit or live, but if you happen to reside in the city of Butte, that amazing Montana landscape is marred by the remnants of an enormous open pit mine. Not only is it an eyesore, but the water that has filled the pit is deadly to any bird that lands there. As a result, a group of people have taken to some ingenious methods to deter birds from landing in the man-made toxic lake for too long.

When they first started, the only tool they had available was a rifle. Scaring birds this way is not the most effective way for all species, though, so lately they have been turning to other tools. One of which is a custom boat built on a foam bodyboard which uses a plethora of 3D printed parts and sensors to allow the operator to remotely pilot the boat on the toxic lake. The team also has a drone to scare birds away, plus an array of other tools like high-powered lasers, propane cannons, and various scopes in order to put together the most effective response to help save wildlife.

While this strategy runs the gamut of the tools most commonly featured here, from 3D printers to drones to lasers, the only thing that’s missing is some automation like we have seen with other drone boat builds we’ve featured in the past. It takes quite a bit of time to continually scare birds off this lake, even through the winter, so every bit of help the team can get could go even further.

Continue reading “Saving Birds With 3D Printed Boats” →

Identifying Creatures That Go Chirp In The Night

September 8, 2020 by Tom Nardi 26 Comments

It’s common knowledge that bats navigate and search for their prey using echolocation, but did you know that the ultrasonic chips made by different species of bats are distinct enough that they can be used for identification? [Tegwyn☠Twmffat] did, which is why he came up with this impressive device capable of cataloging the different bats flying around at night.

Now this might seem like an odd gadget to have, but if you’re in the business of wildlife conservation, it’s not hard to imagine how this sort of capability might be useful. This device could be used to easily estimate the size and diversity of bat populations in a particular area. [Tegwyn☠Twmffat] also mentions that, at least in theory, the core concept should work with other types of noisy critters like rodents or dolphins.

Powered by the NVIDIA Jetson Nano, the unit listens with a high-end ultrasonic microphone for the telltale chirps of bats. These are then processed by the software and compared to a database of samples that [Tegwyn☠Twmffat] personally collected in local nature reserves. In the video after the break, you can also see how he uses a set of house keys jingling as a control to make sure the system is running properly.

As winner of the Train All the Things contest back in April, we’re eager to see how the Intelligent Wildlife Species Detector will fare as the competition heats up in the 2020 Hackaday Prize.

Continue reading “Identifying Creatures That Go Chirp In The Night” →

Approaching The Drop Location: Seeds Away!

September 4, 2020 by Brian McEvoy 8 Comments

Arbor Day is a holiday many countries dedicate to planting trees, but with the steady encroachment of climate change, we need to maximize our time. Dronecoria doesn’t just plant a tree; it sows “hectares in minutes.” A hectare is 10,000 square meters or 2.471 acres. These aren’t the drones you’re looking for if you intend a weekend of gardening, this is in the scope of repopulating a forest with trees or reinvigorating a park with wildflowers. The seed balls in the hopper are 10kg of native seeds coupled with beneficial microorganisms to help the chances of each drop.

The drone’s body is laser cut from what looks like baltic birch plywood. The vector files are available in Illustrator (.ai) and CAD (.dxf) formats released under Creative Commons BY-SA, so give credit if you redistribute or remix it. In the 3D realm, you’ll need a SeedShutter and SeedDisperser, and both models are available in STL format.

We have other non-traditional seed spreading methods like canons, but it is a big job, and if you’ve build something to pitch in, drop us a tip!

Auto-Tracking Sentry Gun Gives Deer A Super Soaking

January 26, 2018 by Dan Maloney 29 Comments

Things rarely go well when humans mix with wildlife. The problems are exacerbated in the suburbs, where bears dine on bird feeders and garbage cans, raccoons take up residence in attics, and coyotes make off with the family cat. And in the suburbs, nuisance wildlife can be an intractable problem because the options for dealing with it are so limited.

Not to be dissuaded in the battle to protect his roses, [dlf.myyta] built this motion-activated sentry gun to apply some watery aversion therapy to marauding deer. Shown in action below against a bipedal co-conspirator, the sentry gun has pretty much what you’d expect under the hood — Raspberry Pi, NoIR camera, a servo for aiming and a solenoid valve to control the water. OpenCV takes care of locating the intruders and swiveling the nozzle to center mass; since the deer are somewhat constrained by a fence, there’s no need to control the nozzle’s elevation. Everything is housed nicely in a plastic ammo can for portability and waterproofing. Any target that stands still for more than three seconds gets a hosing; we assume this is effective, but alas, no snuff films were provided.

We’re not sure if [dlf.myyta]’s code can discern friend from foe, and in this litigious world, hosing the neighbor’s kid could be a catastrophe. Perhaps version 2.0 can include image recognition for target verification.

Continue reading “Auto-Tracking Sentry Gun Gives Deer A Super Soaking” →

Hackaday Prize Entry: Automated Wildlife Recognition

June 14, 2017 by Donald Papp 11 Comments

Trail and wildlife cameras are commonly available nowadays, but the Wild Eye project aims to go beyond simply taking digital snapshots of critters. [Brenda Armour] uses a Raspberry Pi to not only take photos of wildlife who wander into the camera’s field of view, but to also automatically identify and categorize the animals seen using a visual recognition API from IBM via the Node-RED infrastructure. The result is a system that captures an image when motion is detected, sends the image to the visual recognition API, and attempts to identify any wildlife based on the returned data.

The visual recognition isn’t flawless, but a recent proof of concept shows promising results with crows, a cat, and a dog having been successfully identified. Perhaps when the project is ready to move deeper into the woods, elements from these solar-powered networked birdhouses (which also use the Raspberry Pi) could help cut some cords.