During firefighting, rescue missions or inspections in the deep sea, emergency forces often reach their limits. In the future, they will get support in their difficult job: Mobile robots should help them with the help of AI.
A recent report from the platform “Learning Systems”, on which scientists from the Karlsruhe Institute of Technology (KIT) are involved, provides information on the potentials and concrete benefits of AI in this field of application. And moreover, technical and societal challenges, as well as the conditions that must be created for the reliable and cost-effective use of AI in hostile environments.
You might also be interested in this: How a robot saved Notre-Dame
These are the advantages of intelligent, self-learning robots
Especially in dangerous or potentially health-damaging situations and in hard-to-reach terrain their use makes sense. Especially since they often work more economically or even make an assignment possible. But before the application actually works in hostile environments, some technical challenges still need to be overcome. One of them: autonomous learning in unknown areas. In addition, it must be worked out how a collaboration between human and robot can look concrete.
One of those who are convinced of the potential of such mobile helpers is Professor Holger Hanselka, President of KIT: “The use of artificial intelligence is associated with tremendous opportunities for our society, especially in disaster control, the decommissioning of nuclear power plants or in maritime areas great to effectively support professionals with the help of AI. “
Especially with autonomous systems that are to be used in a crisis, IT security is enormously important. “That’s why KIT’s research focuses not only on protecting the external borders of a complex IT system, but also on every single part, and in particular integrating its expertise in IT security into the Learning Systems platform,” says Hanselka.
You might also be interested in this: The generation robot
Whether under fire or under water: AI robots can help people
Two application scenarios created by the Neoliberal Environments workgroup are already demonstrating how Artificial Intelligence can support disaster response and exploration and maintenance missions in about five years. The one illustrated, like AI-based robots of the fire department on the ground and out
to be able to work on the air during the fire of a chemical factory. Creating a detailed picture of the situation, setting up a communication and logistics infrastructure for rescue work, looking for injured persons and identifying and containing sources of danger: all of this can then be the task of these systems with the help of multi-sensor technology.
You might also be interested in this: Robots – we have to talk, but different
And even under water, the mobile robots should have a lot on it. In the future, for example, they can take over the maintenance of the foundations of an offshore wind turbine, navigate independently in the deep sea and take over the planned planning steps. If they can not do one of their own, they ask for assistance from divers or remote-controlled systems.
Ever higher degree of autonomy through machine learning
Jürgen Beyerer, head of the working group Hostile Systems of the platform Lernende Systeme and professor for interactive real-time systems at KIT, explains: “You need to be intelligent and at the same time robust against extreme conditions and under unpredictable conditions to find your way independently. “
Until the AI-based systems can do that, they could also be remotely operated by emergency personnel and the data collected could be used to develop intelligent functions. “Gradually, the systems are becoming increasingly autonomous and can eventually improve themselves through machine learning.”
At the moment, according to the report, they are also working on being able to serve global markets with their self-learning robots. From the development of suitable infrastructures to the creation of standards for business and research and the flexibilization of the procurement market: there is still a lot to do, but Germany is well positioned.