|
As cooperative agents, sensors must be able to communicate in a manner that has real meaning in reference to the goals of the stakeholders. These goals are complex and involve the consideration not only of the squad in question, but other squads, platoons, and enemy forces which may be geographically separate, yet critically important. In order to facilitate this communication, sensors must be designed to operate in the perspective of the soldier, not the technologist.
This means the focus and the value of sensors comes in event recognition rather than signal recognition. Sensors are designed to collect data based on a programmed set of parameters. Simply presenting this (raw) data does not tell the soldier any valuable information as it relates to real goals which are more complex than whether or not something crossed the receptive field of a sensor. Interpreting this data through algorithms into assumptions about the nature of such detections ("was the detection caused by a truck or a group of people") is not the correct way to solve this problem either. Automated assumptions are brittle when used in dynamic, fast-paced adversarial situations, where enemy tactics are constantly adapting to pose new threats. Instead, real advances in sensor utility will come when data collected by these sensors is organized and integrated in ways that make enemy actions apparent. The key is to support rather than supplant the decisions made by the stakeholders. In other words, instead of trying to make assumptions on behalf of the soldier, these sensors should be providing information that makes the decisions easier for the soldiers to make. |