Firewatch: Brazil

Image by Ria Sopala from Pixabay

When human eyes can not fully predict or spot a threatening wildfire in time, artificial intelligence can and will. Meet the innovative system Brazil’s Sintecsys put in place that aims to drastically reduce intervention time for authorities in case of a wildfire.

Wildfires are disastrous, no question about that – the history and chronology of documented wildfires spans widely, the first wildfire known of is documented by the Guinness world book roughly 419 million years ago. Wildfires aren’t inherently very harmful for nature, as with time, those areas are rejuvenated, yet taking global warming into account and how it influences the appearance of wildfires, this becomes a major problem for planet Earth.

In recent times, global warming has influenced the frequency of wildfires around the world greatly, only in 2019 and 2020’s wildfire seasons we’ve seen disastrous outcomes – California alone had a total of 259,823 acres burned in 2019, roughly 8.000 fires erupted in the Pantanal reservation in Brazil in the same year.

The wildfire in Australia was the largest in the last years, reportedly 46 million acres have burned, 3 billion animals were affected, all this spanning across three and a half months. These fires contributed to a high number of carbon dioxide atmosphere pollution, animal and human losses and major property destructions.

In this regard, a new policy for combating wildfires appeared in Brazil, part of the United Nations SDG 2030 efforts. The Sintecsys Brazilian startup came with an innovative solution – artificial intelligence guided cameras installed on towers at high altitudes and overseeing approximately 8.7 million acres of wildlife in Brazil, including the Amazon rainforest.

How does this system work? The installed cameras send datasets of images captured during day and night to a monitoring center where the artificial intelligence’s system uses pre-programmed  labels when scanning the images in order to detect fires (at night) and smoke (during daytime) and if detected, alerting local fire fighting forces. Shortly, the AI takes image sets from cameras, scans them and then sends alerts if smoke or fire’s detected. The success rate of preventing fires and identifying them stands at an approximate 95% after the initial data scans and image processing.

Sintecsys Brazil therefore received the task of monitoring most of the wildlife in Brazil through their newly installed system, with over 50 cameras already deployed in the field (2019 data). This effort contributes to SDG 13 – Climate Action by reducing forest fires and CO2 emissions and SDG 15 – Life on Land by avoiding wildfires and protecting the land.



What do you think?

Sustainable city development: Finland's CitiCAP initiative

Microsoft's sustainability lesson in Minecraft!