Leading cats: How to effectively command collectives

The purpose of this work is to enhance the understanding of command and control in collectives, paying particular attention to coalition situations when multiple, typically autonomous agencies are expected to work together harmoniously. There are two different approaches to dealing with commands, one approach focuses on individual command components and the other studies the interplay between different commands. This paper is a first attempt at bringing these two views together, with the aim of understanding what is needed for human autonomous systems to work as collectives through command and control systems that achieve their full capability through "command by intent", while also encouraging and encompassing new agile behaviour both at the individual and organizational level. We have identified key aspects of commands in collectives (both from existing literature and from our own work) and demonstrate our findings in case studies taken from hurricane Katrina, 1995 Oklahoma city bombings, the 2001 attack on the Pentagon, as well as some incidents of lower complexity, such as an oil spillage on a motorway.