However, the algorithm itself doesn’t provide definitive answers, and a semi-automatic system may not always lead to improved human performance. To address this, the system needs to deliver relevant information to the operator effectively.

To achieve this, we established an integrated sensing and decision support lab. This environment includes real-time interactive virtual simulations, operator-in-the-loop experiments, and distributed data collection setups.

The results show strong evidence that accurate prompts can significantly enhance the effectiveness of drones in detecting military fleet movements. This article will explore the experimental framework, along with the concepts, algorithms, and findings from the tests.

The main challenge addressed here is decision support in counter-insurgency operations, which requires continuous monitoring of large areas—something MTI is well-suited for. However, we also need a sensor capable of clearly identifying potential threats.

While MTI data offers broad coverage, it doesn’t provide clear identification. For this reason, the Predator UAV’s video sensor is a good choice because it can get close to the scene, and the video data it captures is straightforward for operators to interpret.

Unfortunately, the narrow field of view of the Predator’s video sensor makes it difficult to cover large areas effectively. That’s why we aim to combine the strengths of both systems—using MTI to alert the drone camera operator to specific directions of interest.

The cue algorithm plays a crucial role in integrating MTI and Predator video sensors. It analyzes MTI data to detect signs of suspicious activity, such as a fleet, and then alerts the operator accordingly. In this experiment, we used a fleet inspection algorithm developed from MTI radar data collected during the “Silent Hammer” experiment on the Lincoln Multitasking ISR Experimental Platform (LiMIT).

Our goal is to enable Predator drones to effectively survey large areas using their own recognition capabilities. We’re exploring how best to do this. In our experiment, we measured how a Predator drone could perform better in area surveillance compared to a non-prompted system.

The first part of our setup involved ground vehicle simulation. We recreated scenarios involving fleets and riot vehicles. Next, we simulated a Moving Target Indication Radar (MTI) system. MTI is a radar technology that allows for quick wide-area scanning and detection of moving vehicles.

Once the MTI data was collected, it was passed to the cue algorithm. This algorithm automatically analyzed the data and provided indicators of potential activity. For example, if the algorithm was designed for fleet detection, it would highlight a suspicious fleet.