Intelligent Habitat Restoration Under Uncertainty

Authors

  • Tommaso Urli NICTA and the Australian National University
  • Jana Brotánková James Cook University
  • Philip Kilby NICTA and the Australian National University
  • Pascal Van Hentenryck University of Michigan

DOI:

https://doi.org/10.1609/aaai.v30i1.9898

Keywords:

Habitat Restoration Planning, Conservation Planning, Constraint programming, Large neighbourhood search, Uncertainty, Optimisation

Abstract

Conservation is an ethic of sustainable use of natural resources which focuses on the preservation of biodiversity, i.e., the degree of variation of life. Conservation planning seeks to reach this goal by means of deliberate actions, aimed at the protection (or restoration) of biodiversity features. In this paper we present an intelligent system to assist conservation managers in planning habitat restoration actions, with focus on the activities to be carried out in the islands of the Great Barrier Reef (QLD) and the Pilbara (WA) regions of Australia. In particular, we propose a constrained optimisation formulation of the habitat restoration planning (HRP) problem, capturing aspects such as population dynamics and uncertainty. We show that the HRP is NP-hard, and develop a constraint programming (CP) model and a large neighbourhood search (LNS) procedure to generate activity plans under budgeting constraints.

Downloads

Published

2016-03-05

How to Cite

Urli, T., Brotánková, J., Kilby, P., & Van Hentenryck, P. (2016). Intelligent Habitat Restoration Under Uncertainty. Proceedings of the AAAI Conference on Artificial Intelligence, 30(1). https://doi.org/10.1609/aaai.v30i1.9898

Issue

Section

Special Track: Computational Sustainability