Automated Task Allocation Models for Armored Platoons

Author(s):

Mark Dranias, Angelique Scharine, Alfred Yu, Sarah Al-Hussaini, Craig Johnson, Evan Carter, Laura Marusich, Ahmed Khalil, Yoonjae Lee, Efstathios Bakolas, and Gregory Gremillion

Abstract:

Although armored platoon crew members have typically had defined and static roles, the next generation of combat vehicle, with reduced Soldier-to-system ratio, partially manned formations, and increased integration of autonomous agents, will require a more flexible approach to resource allocation and task assignment. Several models have been developed to automate and optimize this task assignment using a variety of approaches. Here we implement two approaches: one that makes assignments empirically, learned from expert demonstrations; and another that optimizes performance by maximizing the value of the assignment using a priori-defined values. Both approaches have their limitations, but we demonstrate how they might work in this future application space and discuss their implications.