Code for our Wildfire Suppression OASYS application and prior published solutions can be found at the two following GitHub repositories:

This repository contains a Cython implementation of Wildfire Suppression, as well as our open-agent Communicative I-POMDP decision process model and corresponding fully online CI-POMCP planning algorithm. This code is based on our UAI’2022 publication: “Decision-Theoretic Planning with Communication in Open Multiagent Systems”

This repository contains a Java implementation of Wildfire Suppression, as well as our approach for solution for scalable reasoning under agent openness that selectively models only a subset of neighbors, then extropolates their behaviors to the collective whole, and reasons using our I-POMCP online MCTS planning algorithm for the open-agent I-POMDP-Lite decision process. This code is based on our AAAI’2020 publication: “Scalable Decision-Theoretic Planning in Open and Typed Multiagent Systems”