Forestry is amongst the industries most impacted by the push towards Artificial Intelligence (AI). Now, global scientists are using AI to predict complex systems, including forest management, animal populations and power grids, identifying disaster’ tipping points.’
“History is full of harmful critical transitions, such as financial market crashes, disease outbreaks and blackouts,” according to Professor Gang Yan, a global expert in network science based at China’s Tongji University, whose research, Early Predictor for the Onset of Critical Transitions in Networked Dynamical Systems, uses machine learning to create two types of AI called neural networks.
And the results speak for themselves.
In one test, scientists used 20 years of real-world data on vegetation and rainfall in Central Africa, training AI to predict precipitation rates. As a result, it claims that AI predicted what had happened to the ecosystem, even when it was only given data for 10% of the nodes to learn from.
Wood Central understands that scientists optimised the first one to understand the functioning of and connections across systems structured like large networks with many nodes: “For example, in an ecosystem, each node would be a geographical location where researchers would collect data about how many animals or trees live. Nodes could also be different parts of the power grid or areas where disease outbreaks occur,” New Science reports.
Scientists also designed a second neural network to analyse how networks change over time. “So, the first network would process data about each node and their interactions, then feed into the second network, which detects patterns in data that recur over time and predicts future tipping points,” New Science added.
According to Professor Yan, past studies have focused on identifying data features that increased or decreased as a tipping point approached. But, he said, the scientist’s AI can further: “It can pinpoint the precise conditions that lead to system collapse,” asserting: ‘Watch out, if the system reaches this [specific] condition, it will collapse immediately.'”
For Dr Karen Abbot, a biology specialist from Case Western Reserve University in Ohio, the new approach is a “powerful” tool that can aid scenario planning. Even if machine learning doesn’t offer the same level of insight into why a tipping point occurs as a full-fledged mathematical model might, it can help.
However, she said AI’s advantage is that it can deal with incomplete or sparse datasets: “We really, really need both. Machine learning is telling us how to get more clues out of data, and theory is telling us what to do with those clues.”
As for the future, Dr Yan wants to apply the model to more systems—such as floods, wildfires, and disease outbreaks—to gain greater insight into not only when the tipping points happen but also why: “We aim to delve deeper into the algorithm’s black-box nature to explicitly uncover the specific features used for predictions.”
Forestry among the top 10 industries impacted by AI
Last year, Wood Central reported that Goldman Sachs ranked forestry amongst the top 10 industries most impacted by machine learning.
According to the report, agriculture, which broadly incorporates farming, fishing, and forestry, will be subject to major disruption. Up to 90% of agricultural workers will be subject to automation to varying degrees.
In September 2018, the World Economic Forum (WEF) reported that “AI could be the game changer for the world’s forests.” At the time, the WEF noted that forest management is a good example of how technology-first approaches can quickly deliver results.
The WEF said predictive analytics and machine learning models are “helping scientists and authorities in different parts of the world in the quest for better forest management—this also includes restoring areas damaged by fires, logging, or clear-cutting.”
- For more information on Early Predictor for the Onset of Critical Transitions in Networked Dynamical Systems, visit the Physical Review X.