Status: 10/30/2022 11:03 am
Critical infrastructures such as water, electricity or traffic are threatened by sabotage or natural disasters. Research and science can contribute to protection, for example with the help of artificial intelligence.
By Annemarie Neumann, SWR
Thanks to the Internet, global trade and global tourism, the world is more connected than ever. But this also makes critical infrastructures such as electricity, transport and the health system more vulnerable.
Just recently, cable sabotage paralyzed the Deutsche Bahn network. It is known from past attacks that the attackers often prepare for a long time. In a hacker attack on a power grid in Ukraine in 2015, the perpetrators had penetrated the system months earlier. To be better prepared for disruptions and hidden attacks, Research and science are therefore also concerned with the security of critical infrastructures.
What is critical infrastructure?
Critical infrastructures are according to the Federal Office for Information Security (BSI) Organizations or institutions that are particularly important to the state community. This includes sectors such as health, transport and energy. Their failure has dramatic consequences, such as long-term supply bottlenecks or disruptions to public safety. In order to keep far-reaching damage as low as possible, they therefore need special protection.
Versatile Threats
The protection of a critical infrastructure depends on the sector and the individual situation. Because with every structure there are different dangers. All areas that work at least partially digitally are at risk from cyber attacks. In the areas of food and water, extreme weather events pose a particular risk for the supply of the population. Deliberate actions or technical and human errors can also affect the sectors to varying degrees. According to the Federal Office for Civil Protection and Disaster Assistance, an all-hazards approach that includes all possible scenarios is therefore taken into account when protecting the individual systems.
Research for more protection against cyber attacks
Research is being carried out in all areas into the protection of critical infrastructures. In this way, possible dangers are analyzed, such as possible extreme weather events in climate research. Protection concepts are developed or the resilience of critical infrastructures is increased, for example against cyber attacks or intentional manipulation by artificial intelligence. “In principle, one can say that a lot of research deals with critical infrastructures without using the keyword,” says the scientist Sadeeb Simon Ottenburger from the Karlsruhe Institute of Technology (KIT).
He deals with the resilience of critical infrastructure. According to the Ministry of the Interior, resilience refers to the “ability of a system to withstand or adapt to events in order to maintain or quickly regain functionality”.
Artificial intelligence as an early warning system
Potential threats to the security of the energy supply are systematically researched at KIT. Artificial intelligence can help to identify and eliminate possible weaknesses in technical systems in good time. With increasing digitization, the various sectors of the critical infrastructure are becoming more interconnected and dependent on one another.
A longer-term power failure could have far-reaching consequences in other areas of critical infrastructure: For example, traffic could be disrupted, there could be bottlenecks in medical care and drug production, or the drinking water supply could collapse. Artificial intelligence can act as an early warning system here. Otherwise, errors in the system could be detected too late or overlooked, especially in highly complex, networked systems.
Simulations should find errors at an early stage
A research group led by computer scientist Eric Veith from the University of Oldenburg is also investigating how artificial intelligence can contribute to the safe and sustainable operation of critical infrastructures. The aim is to increase resilience, i.e. resistance to unforeseen events, using artificial intelligence methods such as independent learning.
In simulations, for example, the network data and communication are tested with the help of algorithms in order to find errors and faults at an early stage. For the time being, according to Veith, research is concentrating on the area of power supply, later this should be applied to all sectors of critical infrastructure.
Various exercise scenarios
In addition to research, each sector also has strategies and legal frameworks to protect critical infrastructures from disruptions and failures. In addition, the federal government, local authorities and individual institutions carry out exercise scenarios and management exercises in order to be able to react better in an emergency.
According to Ottenburger, however, your own provision can also be advantageous, since it may not be possible to help all private individuals equally. Veith makes it clear that everyone has a responsibility to deal with this topic sensitively. At this point, according to Ottenburger, “science should also discuss with the critical infrastructures, politics and society” so that the risk of disruptions can be minimized in the future.