The latest developments at NATO's external borders due to the war in Ukraine are bringing the issue of national and alliance defense more and more into focus. Basically, quick and precise reconnaissance can be crucial in the event of an attack. Therefore, the Bundeswehr, its primary digitalization partner and the software and artificial intelligence (AI) company Helsing are testing a sensor-based system that can use AI to monitor large sections of terrain and accelerate tactical reconnaissance. The project is called MITA – Military Internet of Things for tactical reconnaissance – and has already successfully completed the first two test phases in December 2022 and May 2023.
Precise information for the right decisions
Particularly in the event of an attack, precise information is the basis for military decisions: Where will the attack take place, when, with what forces, from which direction? Today, conventional sightings are predominantly used: soldiers observe the terrain from vehicles, for example, or analyze drone images. The information is usually passed on to the command post by radio. In the command post, the observations are collected, interpreted and entered manually on a situation map. This analog process of information transfer and interpretation requires a lot of time, and transmission errors can easily occur, for example.
Monitor large areas with sensor networks
The interaction of sensors and AI-supported reconnaissance at MITA could significantly accelerate and improve the quality of this process. Situation reports could be created automatically and made significantly more precise with real-time information. Large areas can be permanently monitored using sensors with UWB technology, which BWI developed together with LaterationXYZ. The sensors built into small spheres are inconspicuous, robust and cost-effective. Each ball can detect movements, for example, via a type of light barrier as well as via a ground radar. All sensors are networked with each other and can therefore cover large areas for reconnaissance, including those that are difficult to access or reconnaissance for security reasons. In addition, a link to the Bundeswehr's existing sensor technology is possible. In this way, the information from the Bundeswehr's electro-optical sensors, weapon systems and drones is supplemented by data from sensors installed in the field, so that the threat situation can be determined and assessed in combination. And: Due to the frequencies used, the UWB sensors are difficult to identify by enemy reconnaissance.
3D situational awareness in real time using automated reconnaissance and data fusion through AI
For example, if an enemy tank penetrates the monitored area, the sensors in the field are activated. You can see that an object is moving there and in which direction it is heading. The target area can then be specifically identified using drones. The AI system detects and locates the object and classifies it.
Using software from Helsing, the sensor data is automatically merged, evaluated and converted into a three-dimensional situation picture in real time. Work that would otherwise have to be done manually. This means that the data is available without any delay and can be used by the surgeon to make decisions. The sensor network provides the military leadership with a large amount of independently collected information that is evaluated. The operator selects the weapon system and ammunition and assigns the target. The soldiers at the effector receive the coordinates of the target in real time and can therefore fire at the enemy object without delay. In an operational situation, the system can accelerate the entire process from reconnaissance to command and thus save valuable time, which in case of doubt can mean the difference between life and death.
Claus Hirschmannis head of the MITA innovation experiment at BWI innoX