Information supply during military missions: Relevant or not?
B. Beernink, A.M. Goedegebure, N. Van Kaam, R.W.L. Van der Zon · Research Repository (Delft University of Technology) · 2011
During modern military operations the environment is dynamic and unpredictable. There are many actors in the area of the operation, military as well as non-military, friendly and non-friendly. Combined with personnel reductions this creates the need for a flexible and multifunctional military organisation. In an organisation where less people need to do more it is very important to utilise the skills of every individual. Utilising those skills will create a “power to the edge” structured organisation where decision making is also done at lower levels. This results in groups in the field becoming more autonomous. For autonomous groups to be able to make decisions there is a need to facilitate anticipation and increase efficiency. To reach these goals it is important that a group has a good situational awareness. Providing a person with not only information directly addressed to him, but also with information that was sent to others, could increase situational awareness. Autonomous groups could then be better able to anticipate on future events and make better decisions. However, to prevent information overload a relevance system determines if a message is also relevant by calculating a relevance score for every user. The main research question is how the process of transmitting information can be automated and how this can facilitate anticipation and increase efficiency. This question was researched as a bachelor project for Computer Science at TU Delft. The project was part of the Smart Operations programme at TNO. The execution of the bachelor project was carried out using the situated Cognitive Engineering (sCE) methodology. For a large part this methodology has determined how the project was executed in different phases. In the implemented system, each user can use an application to view the messages that are relevant for him. Latent Semantic Indexing was used for the determination of relevance based on message content, while a fuzzy expert system was used for the overall determination of relevance, which was also based on message metadata. A pilot was conducted with a specialist of the Royal Netherlands Navy. A dynamic scenario was simulated and the system determined the relevance for all messages in the scenario and showed the expert a message list ordered by the relevance. The expert indicated that the developed system has the potential to assist in facilitating anticipation and increase efficiency. He also provided recommendations to improve the current system. The recommendation by the expert was to distinguish separate different message types, for example situational reports and planning messages. Next to this recommendation of the expert we also came to our own recommendations. The first recommendation is to obtain a better dataset for the pilot. The current dataset contained few messages from the environment other than the command center. Furthermore some system components to simulate a scenario are not implemented yet. To do a better test it would be useful to implement a Mission Task Manager, Role Manager and Position Emulator.