Robust static planning tool for military village search missions: model and heuristics

Abstract

In the contemporary military environment, making decisions on how to best utilize resources to accomplish a mission with a set of specified constraints is difficult. A Cordon and Search of a village (a.k.a. village search) is an example of such a mission. Leaders must plan the mission, assigning assets (e.g. soldiers, robots, unmanned aerial vehicles, military working dogs) to accomplish the given task in accordance with orders from higher headquarters. Computer tools can assist these leaders in making decisions, and do so in a manner that will ensure the chosen solution is within mission constraints and is robust against uncertainty in environmental parameters. Currently, no such tools exist at the tactical or operational level to assist decision makers in their planning process and, as a result, individual experience and simplistic data tables are the only tools available. Using robustness concepts, this paper proposes a methodology, a mathematical model, and resource allocation heuristics for static planning of village searches that result in a decision-making tool for military leaders.

Dynamic rescheduling heuristics for military village search environments

“…To the best of the authors’ knowledge, the village search problem has not been previously studied and addressed except in our previous work. 1 There are numerous classes of problems that share concepts with the village search problem such as the knapsack problem, 6 the traveling salesman problem (TSP), and resource allocation problems. In the interest of space, we will only describe research that closely approximates our domain in terms of the modeling framework and the resource allocation heuristics.…”

“…Robustness can have a variety of meanings that are heavily dependent upon the problem domain and the problem solver. A standardized framework for defining the term robustness that results in a quantifiable metric was developed by Ali et al 4 This robustness metric has been adapted to the problem of village search planning and used as the foundation for a mathematical model of the static village search environment by Maxwell et al 1 In this section we provide a brief summary of the robustness procedure and how it is applied to this domain. For a detailed description, the reader is referred to Maxwell et al 1 or Appendix A.…”

“…The Dynamic Genetic Algorithm (DGA) is a modified version of the village search genetic algorithm (VSGA) presented by Maxwell et al 1 It differs from the static VSGA in its construction of its chromosomes, the population size, and the usage of the operators on the chromosomes. Though there are many research efforts such as that by Alcaraz and Maroto 25 that use genetic algorithms in a variety of environments, the operators for the static and DGAs are inspired by the work of Wang et al 26 that demonstrated their efficacy in computing resource allocation environments.…”

“…In our initial work, 1 we described robust resource allocation for a static planning environment where the plan is created a priori offline and resources and time available are not a primary concern. These plans are completed before the search mission begins and no feedback from the mission execution is used to modify the plan.…”

“…When reallocations are required, then fast dynamic algorithms are needed to reallocate resources within the time constraints of an ongoing mission. Adding automated dynamic reallocation components to the Robust People, Animals, and Robots (RoPARs) tool developed by Maxwell et al, 1 or creating a stand-alone solution (independent of RoPARS) for village search replanning in both training and operational environments would greatly assist military leaders in their decision-making process. The stand-alone solution would allow static mission plans to be evaluated for dynamic reallocation regardless of the method used to create the static plan.…”