DARPA's ACTM program

The DARPA AI-assisted Climate Tipping-point Modeling (ACTM)1 program explores the use of third-wave AI methods to enhance the complex interconnected processes modeling climate change.

  1. The main goal is to develop hybrid AI models that combine Artificial Intelligence/Machine Learning (AI/ML) models with conventional physics models of the climate and earth systems that capture missing physical, chemical, or biological processes and identify causal factors with sufficient computational efficiency to explore decadal scale effects and characterize tipping points and bifurcations in these systems.
  2. The secondary goal is to develop new methods to assimilate diverse data into models and estimate "value of new data" to enhance confidence in target-specific forecasts relative to state-of-the-art techniques.

HAIKU Team

BAE Systems2 and AIMdyn, Inc.3 continue their fruitful collaboration applying Koopman to diverse DoD problems (Gamebreaker, COMBAT, and DITTO). Dr. Michael Planer (BAE Systems) is Principal Investigator, leveraging experience in physical and AI systems. Dr. Maria Fonoberova and Prof. Igor Mezic (AIMdyn) contribute Koopman‐based innovations. Prof. Qinghua Ding4 (AIMdyn) is a climate expert, significantly advancing understanding of the relative contributions of internal and external forcing in Arctic climate variability.

1 Elliott, Joshua. "AI-assisted Climate Tipping-point Modeling (ACTM)." DARPA, https://www.darpa.mil/program/ai-assisted-climate-tipping-point-modeling

2 Fast Labs. "Intelligent Autonomous Systems R&D." BAE Systems, https://www.baesystems.com/en/product/autonomy-r-d

3 AIMdyn, https://aimdyn.com/

4 Ding, Qinghua. UCSB, https://www.geog.ucsb.edu/people/faculty/qinghua-ding