Multi-UAV based helicopter landing zone reconnaissance: Information level fusion and decision support

Abstract

This article presents an information fusion and decision-support system for the multi-UAV based landing zone reconnaissance and landing point evaluation in manned-unmanned teaming (MUM-T) helicopter missions. For this, numerous and heterogeneous data from variety of sensors must be gathered, fused and evaluated. However, payload capacity and on-board processing capabilities are often restricted. Thus, the teaming of multiple unmanned aerial vehicles (UAVs) offers a promising way to overcome these limitations and allows to benefit from heterogenous sensor payloads. Furthermore, measurement and sampling processes are never completely reliable. Hence, achieved observations must be interpreted very carefully, especially if the reliability of such functions is relatively low. Thus, the fusion system presented in this paper is based on a Bayesian network to specifically address this problem. Therefore, information needs of the pilots on safe landing zones are determined and required perceptive capabilities are derived. Consequently, reliability estimations of the applied perceptive capabilities are incorporated. Modelling aspects of the evaluation mechanism are explained and implications of incorporated export knowledge are set out. The feasibility of the implemented system is tested in an exemplary rescue mission, outlining the importance of incorporating automation reliability in automated decision-support systems.