Artificial Intelligence: applications and challenges within the disaster management field

Artificial intelligence (AI) is omnipresent in situations where deep learning algorithms and big data can expedite actions required to resolve complex and potentially harmful events. Thus, it is not surprising that AI has become an important tool in disaster management. The applications and contexts in which AI is used across disaster management stages are reviewed and critiqued in this chapter. Particular attention is given to understanding the ethical, technical, and legal issues that constrain wider application of AI now and in the future. Disasters are contextual and spatially bound; consequently, they require an understanding and appreciation of cultural and local constraints to manage what can be a chaotic situation. Understanding how to align these constraints with the learning paradigms and analytical frameworks of AI is critical in the application of AI across all stages of the disaster management cycle. Misalignments create errors, erode trust, and limit the effectiveness of AI, but can be minimized by using development approaches that involve both AI developers and disaster management practitioners and establishing common standards of evaluation in AI deployment. Until a higher level of trust is achieved, the future of AI will be limited to narrow problem-solving, rather than modeling the complex interdependencies that occur across disaster stages to more effectively manage future events.