Virtual exercises for disaster operations

In an emergency, rescue forces must react quickly and with routine. The prerequisite for this is years of training to gain the necessary experience to assess and cope with different hazardous situations. However, not all operations can be practiced realistically. Immersive virtual and mixed reality applications offer an alternative. imsimity GmbH uses artificial intelligence to develop scenarios that are individually tailored to the trainee's skills.

The rushing rescue workers are confronted with a picture of horror: flames are blazing out of a high-rise building, and the fire has already spread to a factory building. Not far away are the burning debris of a crashed airplane. It brings back memories of a day in September 2001, but this scenario is taking place in a small town in the Black Forest. And apart from a handful of scientists, no one has even noticed the devastation. Because at imsimity GmbH in St. Georgen, the fire is only virtual.

Not real, but authentic

Equipped with virtual reality goggles and a rescue chainsaw, one of the test subjects sets out to explore the scene. Wearing protective equipment is not mandatory, but it does increase the sense of reality - and that's what matters most: "Through the interactive training method using a real chainsaw as an interaction device, the user not only understands the learning content cognitively, but also "grasps" and internalizes it physically," explains Martin Zimmermann, Managing Partner of imsimity GmbH. He is the creative mind behind the virtual chaos that rescue workers can face as part of training and education.

People in the fire and rescue service often take a high risk during their mission. Their top priority is to save lives. In order to be able to work safely and without accidents with a rescue chain saw, careful instruction and regular training are essential. The mixed reality simulator ChainSaw VR from imsimity GmbH makes it possible to virtually learn the various basic cutting techniques on materials such as wood, laminated glass, plastic or metal as well as to act out, emotionally experience and memorize relevant emergency situations. The advantage: The simulation requires neither fuel nor real material for training.

Training programs created with AI

Two neural networks have been integrated into the program. They divide the user into experience levels and automatically generate training plans to provide personalized training. The first neural network was trained with execution data from different experience levels and assigns one of three experience levels (beginner, advanced, and expert). The second neural network was also trained with different execution data and uses collaborative filtering methods to create a training plan based on the collected execution data.

The resulting recommendation system allows the user to directly monitor the success of the executed content and provides information about the quality of the cutting performance of the rescue chainsaw. In addition, the generated training plan addresses the exercises that can further improve the user's execution quality in the future.

In the ChainSaw VR learning scenarios, emergency personnel go through different emergencies with the real rescue chainsaw in hand. The complete immersion in the digital rescue world (immersion) makes the exercise true to reality. The user experiences interactively and in real time the extreme boundary conditions such as darkness, heavy smoke development and the associated limited visibility. Training with an AI application such as ChainSaw VR is suitable for various areas of professional, rescue-related training and further education - for example, for members of the German Federal Agency for Technical Relief (THW), all fire departments or other blue-light organizations.

Application facts


Technology field
Virtual and Augmented Reality
Application industry
Cross-Sector
Other Services
Administration and Security
Field of application
  • Miscellaneous
Value-added activity
Research and Development [R&D]
Miscellaneous
AI Developers

imsimity GmbH

SME
Website