Plot-twist: Could AI save the planet?
By Apala Bhattacharya
AI is helping countries track pollution in oceans, forests and the skies.
Image: Mark Anderson – CC BY 2.0
From building a ‘Facebook for fish’ to helping Chinese cities track air quality, artificial intelligence is helping keep tabs on pollution. Countries are now able to collect more accurate data and predict patterns in forests, oceans and the atmosphere.
Here are four ways AI could be the heroes of our planet’s future.
Predicting pollution
Beijing is using a machine learning tool to monitor and predict air pollution patterns across the city. Green Horizon uses sensors that report data in real time and compiles this with weather and air flow patterns to produce trends in air pollution across the city.
This allows public officials to redirect traffic or even halt industrial production in areas that are overly polluted for a certain period of time when air quality is extremely harmful. It learns from historical data to predict pollution patterns down to a square kilometre up to 10 days in advance.
This could help public officials plan ahead and initiate health precautions if necessary. Two other Chinese cities - Baoding and Zhangjiakou - have plans to use this approach.
Building ‘Facebook for Fish’
A NASA algorithm originally built to track stars is now helping spot endangered fish in the ocean.
Marine biologist Douglas McCauley is using it to track giant sea bass, a species that is nearly extinct. The astronomy algorithm tracks the fish and their migratory patterns by identifying unique patterns on their bodies - just as it does for stars.
McCauley has created a database of the 500 odd giant sea bass left in the world today and has catalogued 115 of them. As a result, he has found new locations where these fish often travel to and where fishing should be banned or restricted.
Tracking these fish allows conservationists like McCauley to better argue for the expansion of marine protected areas that only make up 3.7% of oceans today.
Tracking illegal fishing
Indonesia is the first country in the world to make data on fishing activities public. Ocean fishing usually occurs outside of tight monitoring and regulations, and can often lead to unregulated and illegal fishing that deplete resources and destroy marine biodiversity.
The Minister of Fisheries and Marine Affairs Susi Pudjiastuti has taken to blowing up illegal fishing vessels in the coast of Indonesia. Sharing the data publically, the Minister hopes that increased transparency can lead to better regulation of the fishing economy and save Indonesia several millions of dollars.“We are also cracking down on illegal, unreported and unregulated (IUU) fishing by publishing [and] socialising it, so we get full public support and we can crack down quite successfully,” notes the Minister.
Global Fishing Watch, a website that tracks fishing boats, is now looking to deploy a nighttime imaging system that tracks lights used by fishing vessels at night to map illegal fishing. The technology will fill in gaps in the data and can be used to better prosecute individuals caught fishing illegally.
Sharing vessel data publicly puts international pressure on countries with boats caught illegally fishing. It can also help create policies to protect fragile marine ecosystems from overfishing.
Cutting carbon gases
In the Malaysian state of Sabah, an AI powered plane maps the ecological biodiversity of rainforests. The plane collects data on carbon and nitrogen levels in the air, using these to map out specific tree species.
This also helps identifies land areas with the largest biomass of atmospheric carbon. Protecting these areas can help slow climate change, as they help absorb excess carbon gases from the atmosphere.
This new information will help the Sabah Forestry Department and the Southeast Asia Rainforest Research Partnership mark areas for conservation.
Maybe instead of destroying the world like many apocalyptic movies predict, AI might just help us save it.