Advertisement
AI in Southeast Asia: Machine learning offers new solutions to age-old environmental problems
Artificial intelligence is making its presence felt in the push for sustainability. In the latest of a regular series examining AI's development in the region,CNA examines three tech-driven projects with ambitions to help with climate disasters, biodiversity monitoring and plastic waste.
New: You can now listen to articles. 
Sorry, the audio is unavailable right now. Please try again later.
This audio is AI-generated.
Jack Board
BANGKOK: Imagine a forecast that could provide precise temperature or wind conditions down to a city block level. Or how a storm might inundate an individual housing cluster, giving residents time to prepare for specific, local impacts.
That is what Mr Alex Levy did as the co-founder of Atmo AI sought to develop a technology that could revolutionise the way weather is predicted and communicated, especially with climate change threatening more unpredictability and damage.
“Predicting the weather better is one of the best things that we can do to make the challenges of climate change less painful than they otherwise would be,” he told CNAfrom the company’s headquarters in California.
“Bad events are bound to happen. But those bad events are compounded when they’re surprises and when they come as a shock to the people that experience them.”
After extensive smaller trials, his artificial intelligence-driven platform is on the brink of being rolled out at a national scale for the first time.
SUPERCHARGING WEATHER FORECASTING
It is an example of the fast-developing world of AI colliding with regional efforts to combat climate change and make gains in sustainability research fields.
Atmo AI has signed a contract with the Philippines’ national meteorological agency PAGASA – the Philippine Atmospheric, Geophysical and Astronomical Services Administration – to complement the country’s existing weather forecasting with a machine learning approach.
The promise is faster autonomous forecasts with greater accuracy – potentially a major boon for the agency, considering the Philippines’ geographical reality.
The country is the most typhoon-prone on the planet, enduring about a quarter of the global tally every year. And because of climate change, the stormsare expected to hit harder and move along paths that are difficult to predict and prepare for.
“The question that we set out to ask was, could we take everything that's been learnt in AI and physics, bring it over to atmospheric science and turn up the accuracy, turn up the detail and crank down the cost?” Mr Levy said.
“What if you could give the Philippines better forecasts than America? That was it.”
Instead of solely relying on expensive supercomputers and satellites monopolised by advanced economies like the United States, United Kingdom and Japan, the Philippines will now be armed with its own customised AI tool, which runs on a subscription model with Atmo.
Typical existing weather forecasting worldwide has not changed for about 70 years. While improvements have slowly been made, the tools are fundamentally similar – essentially usingmathematics and current conditions to predict the future.
According to Mr Levy, thishas left many countries with substandard weather forecasting.
“What you've seen is a lot of these global models have gotten progressively better at a slow rate for the area that sponsors them, but are actually very poor in the tropics and subtropical regions,” he said.
That could soon change, from large urban centres like Manila to rural provinces where understanding the weather is important for agriculture. The detail in forecasting could be enhanced by 100 times, Mr Levy said.
“The previous gold standard forecasts would have treated the whole city (of Manila) as four zones. We treat it as 10,000 zones. It’s going to be like putting on glasses,” he said.
“When it comes to things like inundation and rain and flooding, there are significant differences. With much greater detail, and much greater accuracy that empowers you to take actions to protect yourself.”
At PAGASA, the project lead is confident the technology from Atmo AI can help as the met agency shifts focus towards “impact-based weather forecasting”, which measures potential damage from extreme events rather than basic metrics, such as the amount of rainfall. In time, that could help develop an improved early warning system for communities.
Ms Maria Cristina Uson explained that traditional modelling can accurately forecast the nexttwo to three days in a rough three-hour process. Atmo AI promises 14-day accuracy in just 15 minutes.
While she places the cost at 50 million Philippine pesos (US$890,000) per year, Ms Uson believes it will pay off.
“Even though it is expensive, if it is so effective, the benefits you get will be worth it,” she said.
“If we can avoid incurring severe damage caused by storm surges related to disasters such as tropical cyclones because we have this kind of technology, you will no longer mind the expense. You’ll only think about the lives that were saved.”
For now, the Manila project is the only one of its kind in Asia. But other countries in the region are already in various stages of using AI for weather forecasting.
Bangkok is cooperating with Japanese firm Weathernews Inc to better track rain in the city – giving accurate three-hour warnings about downpours. Meanwhile, Meteorological Service Singapore (MSS) has deployed a supercomputer to improve the quality of its forecasting.
Mr Levy told CNA there is scope and interest to extend Atmo AI’stechnology across the region; the company has also previously done test casing in Singapore.
“What made this a great idea for the Philippines makes it a great case for its neighbours. So I expect that will be the case,” he said.
“And I hope that it will spread throughout the world and in doing so will counteract many of these very painful trends that have emerged in the last couple of decades.”
Related:
AI in Southeast Asia: As new frontier opens in scams and cyberbullying, authorities wage high-tech battle
AI in Southeast Asia: As rules are drafted, workers share their worries and wishes
ALL EARS ON HEARING BIODIVERSITY
From radars to rainforests, AI is also finding use in the depths of some of Southeast Asia’s most ancient landscapes.
Students Clara Hernblom and Johan Närvä from the Swedish University of Agricultural Sciences are conducting research in the pristine – and also more degraded – forests of Sabah on Malaysia’s Borneo island.
They aimto better understand the levels of biodiversity and wildlife activity across different landscapes, including restoration sites. The findings could provide insights into the effectiveness of carbon credits, where companies can offset their carbon footprint through the restoration or preservation of forests.
To understand the ecosystem, they have placed dozens of audio recorders throughout the landscapes that they are studying. Over 10-day periods, the devices capture the sounds of the animals, birds and reptiles living there.
Instead of manually having to listen to thousands of hours of recordings and trying to determine species and behaviours, the students can upload their recordings to an AI-powered platform called Arbimon, which can provide analysis in a fraction of the time.
“It seems really promising. This project would not be possible without it. And going forward, we will only get more species that we can identify.” Mr Närvä said.
It is a method being increasingly used around the region and the world – the online tool is free for scientists and researchers to access.
Arbimon originally started as a cloud-based programme for storing and analysing audio recordings at a university in Puerto Rico. Early applications included helping park rangers detect illegal activity, like chainsaw use in the forests of Palawan in the Philippines.
Its capabilities are now much more sophisticated.
The technology uses pattern matching and clustering to discover specific sound signatures that are indicative of present species. It has the potential to give insights into what is happening in the forest – in near-real time.
“For a scientist that has no access to technology, it would take on average 10 to 15 minutes to process one single recording,” said Mr Bourhan Yassin, the CEO of Rainforest Connection, which runs Arbimon. The non-profit organisation is dedicated to protecting threatened ecosystems.
“So think about that and multiply that over 100,000 recordings, which is not a lot for a dissertation or thesis. It would equate to somewhere around four and a half months to get a really good set of detections for a single species.
“With an AI model, you can do this in seconds. Literally, you can process a million recordings in a matter of seconds,” he said.
Mr Yassin said Arbimon is helping to bridge a huge gap between science and research and conservation on the ground. He shared that two to three million recordings are uploaded every week to the platform. Close to two billion analyses have been done so far in about 6,000 projects in almost 120 different countries.
Despite the vast potential for eco-acoustics to be used, especially in rainforest-dense Malaysia and Indonesia, limitations remain around its use in Southeast Asia.
Mr Yassin admitted that governments are still wary about data collection, particularly in national parks, and the hosting of that data outside their own countries.
“The government regulations haven't caught up yet,” he said. “In order for AI to be widely adopted in Southeast Asia, there has to be decentralisation and coming to terms with the fact that AI doesn't have to run entirely locally and cannot just be controlled entirely within the country.
“For these systems, especially systems like Arbimon, that are not serving just one country, they have to be global.”
WHAT TYPE OF PLASTIC IS THAT?
The Philippines has a bad reputation when it comes to plastic pollution. The country ranks as the world's leading polluter, accounting for 36 per cent of the global total plastic waste emissions into the ocean, according to a 2023 UK report.
But for those trying to understand the plastic problem in the region, these were hard numbers to quantify with hard facts.
“The Philippines has been tagged as one of the significant contributors of plastics pollution. But we always go back to the fact that there is no hard data on this,” said Dr Deo Florence Onda, a microbial oceanographer from the University of the Philippines.
“There is no baseline data coming from the Philippines. And many of the classifications were mainly based on statistical inference,” he said.
For Dr Onda, developing countries have been forced to shoulder most of the responsibility and blame for plastic pollution, when it should be divided more fairly, including among plastic producers.
Since 2022, he has been spearheading Plasticount Pilipinas – an effort to create a comprehensive baseline on where plastic is located around the country, what type it is and how much there is. He believes the science-based approach can inform better decision making about the problem.
But counting plastic in a country with the world's fifth longest coastline was a daunting task. Conventional, manual methods would be “very toxic and very intense”, Dr Onda said.
“The idea now is that technologies can actually help us ramp up our capacity to do research, and that's our entry point for AI and using machine learning, coupled with drone technology, as well as higher resolution imaging systems, to actually help us count microplastics and microplastics,” he said.
To tackle the formidabletask of collecting nationwide plastic data, a team of researchers latched onto existing AI technology developed in Japan to collate, synthesise and visualise the problem.
They first had to adapt the AI’s capabilities to a local context – the plastic problem in the Philippines was very different to Japan’s, explained Mr Paul Samuel Ignacio, a mathematics professor at the University of the Philippines Baguio and Plasticount team member.
“AI is very hot right now and everyone wants to do it. But then everyone depends on data, right? So you need image data to train an AI or you won't be able to get a good model,” he said.
Slowly, the team has built up a library of about 50,000 open-source, field and laboratory-generated images to train their machine to identify 16 types of plastic pollution.
By flying a drone over a coastline, for example, the AI can now automatically detect and classify the objects in real time. It can do the same with photos taken by the public, researchers or authorities.
Also in the high-tech crosshairs – microplastics, which are largely invisible to the human eye and are of growing concern. A different application of the AI combined with fluorescent dyes is able to detect and count them.
Average accuracy for the classification of plastic objects is up to about 85 per cent, Mr Ignacio said.
The Plasticount team is contextualising the plastic problem at a local level so different locations can adapt tailoredpolicies, rather than relying on cookie-cutter national plans.
For example, they have learnt that Manila is most plagued by single-use plastics associated with the food and drinks sector, while Palawan province suffers most from fishing industry pollution.
Having a growing set of baseline data will be critical at the negotiating table with problem industries in the Philippines and wider region, Dr Onda said.
“If you want to have a policy that really works, then you need to target the most problematic plastic type. And the data gives you that information,” he said.
Plasticount is working to provide tailored summaries to help the government target what kind of industries to talk to or what kind of plastic products should actually be prioritised.
The Philippines is not alone in the region in using AI to supercharge its plastic data collection – similar projects are underway in Indonesia and Singapore.
Given the regional nature of the plastic problem, the Plasticount team makes its data publicly accessible, with the aim of it becoming a hub for researchers anywhere. In the future, the team hopes harmonised policymaking could help solve pollution issues allacross Southeast Asia.
“Even if we solve the problem in our country, our neighbours do not, we will still be affected by it,” Dr Onda said.
“But this data gives us a deeper understanding of where the plastic is coming from and how to potentially deal with it better.”
Source: CNA/jb(ws)
Related Topics
Southeast Asia Philippines artificial intelligence environmental sustainability Climate change weather
Advertisement
Also worth reading
Content is loading…
Advertisement
Expand to read the full story Get bite-sized news via a new
cards interface. Give it a try. Click here to return to FAST Tap here to return to FAST FAST 
