Name: How A.I. could save Hawaii's rainforests
Uploaded: 2022-04-25
Duration: 8 min 47 s
Description: A.I. is helping target invasive species that are destroying Hawaii's biodiversity.

A.I. is helping target invasive species that are destroying Hawaii's biodiversity.

How A.I. could save Hawaii's rainforests

Tech

A.I. is helping target invasives species that are destroying Hawaii's biodiversity.

Hawaii's native forests are incredibly unique. They've evolved in isolation over Millennia. Over 90% of Hawaii's plants and animals that are native are actually endemic and found nowhere else in the world. But they're vulnerable. They're sensitive and they're really delicate. And as a result, invasive species just take over and just really proliferate and easily outmaneuver these native species. At this point, there's only the very top reaches of the mountains in some coastal areas that are really housing are native ecosystems. Restoring the balance is absolutely achievable. We're in the process, I'd say of ai data informing our strategy, The nature Conservancy preserve is 8,951 acres and it's part of a larger watershed that's 100,000 acres that has multiple partners. This forest, as you can see here, all of the layers are here, there's nothing that's dominant. It's got ferns and vines and brian fights on the ground, mosses and stuff and then there's a shrub layer, a mid canopy and then a high canopy that is what a healthy forest should look like. Our Hawaiian rainforests need the brioche fights and the mosses to collect water and to percolate water into the ground water is such a valuable resource. We drink the surface water that's collected from this area and when you have invasive plants like ginger, the water just runs off and it doesn't go into the ground, invasive species are really threatening the health of forests in a major way in hawaii invasive species come to the islands. Um perhaps one every five days comes to the island because the biosecurity is really hard. We have ships and planes and cargo coming here. Once those invasive species get established in a native forest, they can just wreak havoc. Manpower on the ground is really hard, especially because we have these areas that are really remote where we have to fly into and the forest is really thick and for people to crawl around in that forest looking for ginger and killing ginger. It's really time consuming. It's really costly. It's really dangerous because we're out there with no communication as we all know. AI has been around for a long time. We use it every day whether we want to or not through our phones and our computers and our Alexis and such. But Ai for conservation is new and we're just starting to realize what it can do for us, including streamlining data processes, helping us find plants with remote means, helping us analyze data and spit out reports and tell us where to go to find some of these plants. Some of the ways we're looking at capturing data and what types of data that apply for this particular project starts with something that's relatively easily available and that satellite imagery and that feeds us a good raw starting point for this project. We've looked at imagery from drones that have been flown around the preserve. We've also looked at helicopter imagery. We're also getting into some on the ground information about where is this particular ginger spotted. So you start looking at a variety of different data sources, all of which becomes signals that we use ai to apply um and understand and predict whether there's ginger in a frame or not. We picked a million ginger because of where the problem emerged. It can still be spotted with the naked eye but it's still confusing for ai. So we wanted to test how good is this ai to really detect something on. Just a visual cue. We have an Ai algorithm that can put little boxes around suspected ginger plants. And so the next step is to give those data points to a crew member and they go out and take care of the weed in the forest. Okay, so there's a good example right here of a big patch of ginger and typically will kill this using herbicide. But what we'll do is cut the big socks off and we'll poke holes in the corm, we have these little drip bottles of herbicide, it's called mezzo Furin muscle and we just put a little drop on the tops of our cut pieces and into the areas on the corn where we've cut and that will kill the the route and hopefully we won't have to come back for a long time and hopefully we get it all, there's really a lot that's at risk. The water is first and foremost. Native forest is where our water comes from. But in addition to water, the native forest is really important for culture. The Hawaiian culture only exists because of our natural resources, cultural resources and natural resources are one and the same. And it's where the native people were the Hawaiians. We'll go to conduct cultural practices, gather for lei making and really connect with their ancestors. So it's really important in that sense. Mm hmm. I think the vulnerability here in hawaii has largely to do with this finely tuned balance of nature that has formed over, you know, countless millennia as someone who identifies as both the native Hawaiian and a scientist. I don't view traditional ecological knowledge and modern technology as being exclusive of each other. Rather I feel they're really vital to inform each other and help us find the best solution. Mhm mhm. In Hawaiian proverb, we say that the land and all that feeds us is chief to whom we are servants And so we really view these creatures these forms of life as where were descended from and because of that they deserve to be treated as sacred and deserve to be treated and cared for like family. The one anchor point of ai when it comes to conservation work is it has to be people first, it has to be anchored on people on the ground that understand the people, the needs, the culture, the nuances that are not just science driven or not just anchored on the ecology. But the triangle got people science and nature and that really is where we have to fit in. We have to fit a I into that construct. We're just starting to scratch the surface of what AI can do for us in the world of conservation because it's not an easy problem. It's really tough and you need the brightest minds, you need industry and business and various sectors coming together to say we want to do something that will help the planet and that's what AI can do for conservation. The biggest optimism here is we're seeing a I embraced the role of science and empower the scientists and conservationists to do this work much quicker and faster and removing a bottleneck that typically would have been a data or a computer or a machine learning expert. We believe we have a very strong pipeline of additional things we want to incorporate to continue building towards the models that we've generated with Ai. I share that optimism if we can continue to control the threat and use innovative technologies and means that allow us to be more effective, more conservative with our resources and being able to reallocate resources to tackle those new threats when they come up, then we have a huge chance and it's achievable