Robots Weekly ?: Outpost Edition ?
We dig the outdoors here at Blue Ion, it’s why we created Outpost. (Check out the podcast if you haven’t already.) So, naturally, I’m always interested to see how AI and nature overlap and interact. Thankfully there is a lot of cool work being done to scratch this itch. Let’s check some of it out! ?
Ever been wandering the woods and get worried that plant you just brushed against might cause some itching? Or hiking while hungry and want to make sure that mushroom isn’t of the deadly variety? Or glimpse a striking bird and wonder what species it is? Species-Identifying Artificial Intelligence got you covered!
The improvement of computer vision and image recognition has led to a growing category of apps that can identify nature using your phone’s camera. Are they perfect? Nope. Are they cool? Yup. And they are far from the only cool thing happening in Nature AI.
What Else Is AI Being Used For? ?
The Elephant Listening Project is now using AI to analyze their trove of audio to recognize forest elephants. This allows conservationists to better estimate population numbers and track the animals through unique voice signatures.
The ELP is also using AI to pick out illegal poaching noises in the audio, like gunshots, so they can alert authorities. And apparently they’re using image recognition to flag online ads for poaching related products. Pretty rad.
ATLAN Space is doing some awesome work in the conservation space. They’re rolling out a pilot project in the Seychelles using AI-powered drones.
The drones will:
- Determine what type of vessel a boat is
- Determine whether the boat is in a protected marine area
- Determine if it is an authorized fishing vessel
- If the vessel hits the illegal jackpot the drone will:
- Note the vessel’s location
- Count the number of people onboard
- Note the ID number
- Notify authorities
Sounds like the Coast Guard is about to get some team members.
The use of cameras on fishing vessels to monitor fishing practices and general rule following has been gaining steam, at least in the Northeast. The downside? There is a lot of footage that needs to be reviewed. The solution? This is a Robots Weekly post so the answer should be obvious.
The Nature Conservancy ran a competition to see if computer vision might be the answer. Spoiler alert, seems like it.
“Winners got close to 100 percent in count and 75 percent accurate on identifying species,” he says. “We proved that automated review is now. Not in five years. And now all of the video-review companies are investing in machine learning.”
There are a lot of cool ways AI is being integrated into farming that will hopefully lead to more sustainable practices, but let’s use wine as an example here. ?
The University of Melbourne tested the use of drones and AI to help with irrigation management and found the combo to be highly efficient and accurate. The algorithms categorized plants by water stress level, flagged the presence of pests and disease, and also pointed out if they were dead. How efficient was the system?
This proves how technology allows a 45-hectare land to be surveyed in 15 minutes and have the data ready a day later.
That’s roughly 111 acres.
One of the co-founders of Wild Me built an AI system that identifies whale sharks based on their unique dot patterns. Where do you get training data for something like this? Space! Turns out whale shark patterning is rather similar to constellations, so he got the help of an astrophysicist and repurposed a star mapping tool. Duh.