Fraunhofer FKIE's measurement campaign for detecting gunshots and drones

The Ukraine conflict confirms that wars are also fought in urban environments. Russia is mainly aiming its attacks at the neighbouring country’s large city centres.

Reliable, technology-supported reconnaissance is a key factor in gaining the upper hand here. Sensors are an important resource for this.

Classifying gunshots and drones

Researchers at the Fraunhofer Institute for Communication, Information Processing, and Ergonomics FKIE are looking into the ways that the systems can provide support with recognising, classifying, and pinpointing gunshots and drones, as well as important information to get an overview of the situation.

At a multinational experiment conducted in Switzerland, the scientists had a week-long opportunity to put their latest research to the test in realistic attack scenarios.

Multinational experiment

Research teams from the NATO member states of the US, France, Hungary, and the Czech Republic were all in attendance

The exercise took place at the invitation of the Swiss Federal Office for Defence Procurement (armasuisse) in Walenstadt in the south of the Canton of St. Gallen as part of the NATO SET-286 research group “Acoustic and Seismic Sensing of Threats in Urban Environments” collaboration.

As well as Switzerland, which is not a member of NATO but is part of the “Partnership for Peace (PfP),” and Germany, represented by Fraunhofer FKIE and supported by the German Armed Forces Technical centre for Weapons and Ammunition (WTD 91), research teams from the NATO member states of the US, France, Hungary, and the Czech Republic were all in attendance.

Using sensors for a better overview

“Urban environments pose a great challenge for detection, reconnaissance, and situational awareness,” explains FKIE research group leader, Dr. Marc Oispuu, adding “The NATO SET-286 research group aims to develop and test innovative technologies that detect dangers in this environment. The focus is on acoustic and seismic sensors. This is because signals, like those emitted by gunshots, explosions, or ground and air vehicles, can be recorded using these sensors without seeing them.”

The data detected can ultimately be entered into a surveillance system and used to provide key support when making decisions.

Conducting exercises with the highest level of security

To make this as realistic as possible, a range of different resources were used in the exercise

'Äuli', a model city in the Walenstadt military training area containing realistic reproductions of houses, gas stations, and supermarkets, provides the ideal setting for one of the largest exercises carried out in terms of the degree of organisation and security required.

The research team’s task was to use the variety of sensor technology located strategically around the model city to detect and classify around 2,000 shots and localise the shooters from a total of eight types of weapons, rifles, and handguns, with and without silencers, indoors, outdoors and from house to house as well as explosions in bunkers and the surrounding area, attacking drones and approaching military vehicles. To make this as realistic as possible, a range of different resources were used in the exercise scenarios.

Valuable data set for research

“As researchers, we do not get many opportunities like this,” concludes Oispuu, adding “We were able to bring back 3.4 terabytes of data. Such a high-scale, diverse, and well-documented data set is immensely valuable when it comes to continuing our research, so we can use this to make a real start on our work."

Measurement campaign

Dr. Marc Oispuu adds, "The data were analysed before being shared and compared with that of other nations. This increased the data set many times over as a result.”

In the fall of 2024, Oispuu and his team plan to put the results they compiled to the test once more in a measurement campaign at the Lehnin military training area in Brandenburg.