can affect the visibility of smoke in images. By continuing to improve ML algorithms with more data, AI-enabled camera detection can reduce false positives and improve the accuracy of smoke detection. However, a key restriction remains in that even AI-enabled cameras typically are not able to see what's happening under the tree canopy. Their capabilities are limited to detecting smoke plumes only when they rise above the canopy. This is a significant limitation as most human-induced fires start at the forest floor, and smoke only breaches the canopy once the fire underneath has had the time to grow quite large.

The process can take up to several hours from ignition, especially if the fire starts as a smouldering fire, e.g., as a result of a discarded cigarette. The delay in detection can mean that by the time fire crews arrive on the scene, they are facing a dangerous job trying to contain the fire. While infrared technology could help to complement the shortcomings of optical cameras, the resolution of these camera systems is typically too low to provide usable images for detecting fires at a great distance.

Training Gas Sensors with Machine Learning

Another approach is to use gas sensors to detect wildfires. Gas sensors are small, wireless devices attached to trees throughout the forest that can “smell” a fire, akin to a “digital nose.” Once smoke is detected, the device sends a signal across the network to alert the authorities. One of the main benefits of gas sensors is that they can be embedded in the forest and can detect fires below the canopy layer while the fire is still in its infancy, allowing for quicker and more effective response and enabling fire fighters to extinguish a fire before it spreads out of control.

However, the sensor-based approach to detecting wildfires also comes with its own set of challenges. To accurately “smell” smoke, the devices are using machine learning (AI) models trained with data from fires and clean air taken from the forest environment. The challenge involves training the models to distinguish between the “smell” of a fire and other ambient gases. For example, the smell of a forest can vary depending on factors such as the type of trees present, the time of day, and even the season. Collecting a broad variety of data to provide a reliable machine learning model can be a tedious challenge. Yet, by incorporating these variables into the training process, the AI models can become more resilient to false positives and more accurate in detecting actual fires. They can even be trained for a specific forest.

To train the machine learning models, researchers create artificial environments in which they burn materials from target forests. The smoke generated from these controlled burns is then fed to machine learning models to teach them what a fire actually “smells” like. This process is repeated hundreds of times to improve the accuracy of the models. The more diverse the training data, the better the AI becomes at distinguishing between real fires and false positives.  

Predictive Capabilities of Satellites

Another approach to wildfire detection is the use of satellites. Satellites have a great overview from above and can use cameras and infrared sensors to detect hot spots and wildfires from space. With the help of AI and image recognition techniques, this process can be automated, and authorities alerted when a fire is detected.

However, one of the key drawbacks with regard to wildfire detection is the resolution and the frequency of updates. Satellites used for detecting wildfires can be either geostationary (about 32,000 kilometers from Earth) or low-orbiting (about 600 kilometers from Earth). One pixel of a geostationary satellite image could cover as much as 500 square meters, meaning a fire would have to be very big to show up at all. And once fires are very big, they become incredibly difficult to extinguish. Low orbiting satellites, on the other hand, are closer to earth and can provide higher resolution (e.g., 100 square meters), but as the Earth rotates below, these low orbiting satellites can provide updates only every six hours for a given spot on earth. This can be mitigated by launching hundreds of satellites to get to frequent revisit times, but that would be quite costly given the relatively short lifespan of low orbiting satellites.  

What satellites excel at, however, is being able to predict the development and spread of wildfires by considering various factors such as terrain, wind direction, and speed. AI and machine learning can be an immense help in predicting the development of wildfires by taking into account huge amounts of data to quickly build accurate models. This information can be passed along to firefighting and evacuation teams on the ground to help them coordinate efforts.

Each technical approach for containing the increasing threat of wildfires presents its own advantages and disadvantages. However, by combining and integrating the information from various detection methods, the advantages of one approach can cancel out the disadvantages of another. AI is a common theme across all solutions and will be able to help coordinate response efforts in real time. Data from gas sensors, cameras, and satellite imagery could be analyzed together to provide a comprehensive view of the wildfire. We would know where the fire started, its estimated size, and its likely path and spread.

By utilizing AI to enhance wildfire detection, we can significantly lower wildfire risks by enabling us to detect and extinguish them in their early stages, before they have a chance to spread out of control. Extinguishing a small fire requires dramatically fewer resources than trying to contain a megafire.