Security Assistance Manager for the Smart Home

SAM Smart stands for "SicherheitsAssistenzManager" (Safety Assistance Manager), which is designed to assist consumers with "tedious" and "error-prone" safety tasks without restricting their autonomy. The final product should proactively warn of possible safety risks in an unobtrusive but clearly visible way - and interact with consumers in a natural way using speech. Its ultimate aim is to retrieve detailed, comprehensible information on the potential threat.

The project seeks to investigate and develop cost-effective security solutions and automated AI-based corrective measures for low-quality products for the mass market. For this purpose, the project relies on a living lab with private households as well as a smart home test lab to be able to research and develop solution approaches. Important parts are the development of a privacy dashboard and an ambient security display, which provides users with an overview of the security of the devices used. Furthermore, a voice-based assistance system will be used. This should enable consumers to assess smart home devices according to their specific risks and to make decisions regarding configuration and further use based on this (e.g. secure router configuration via a device). Furthermore, data protection-friendly solutions for voice assistance systems will be evaluated.

The project is done with close collaboration between the Institute for IT Security (Prof. Eisenbarth and Prof. Mohammadi) and the Institute for Medical Informatics (Prof. Marcin Grzegorzek). The focus of this collaboration is to consider privacy, data protection and data security aspects in all developments of the Institute of Medical Informatics in the areas of pattern recognition and machine learning. Among other things, it is important to rule out the possibility that increasingly complex methods of pattern recognition, which allow for inferences at ever higher semantic levels, can be used to identify individuals even in the case of data sets that are avoidably pseudonymized and anonymized. This work is particularly relevant for health data, where re-identification could lead to exposure of an individuals sensitive medical information. Quantifying and mitigating this risk helps in conducting ethical research and ensuring privacy protection.

3.23 million Euros via the Federal Ministry of Education and Research (BMBF) of the Federal Republic of Germany together with the partners: University of Siegen, open.INC GmbH, nuspace GmbH, automITe-Engineering GmbH, Langlauf Security Automation GmbH. The project is funded by the European Union - Next Generation EU. Grant number: 16KISA074

Project duration: December 2022 - December 2025

News article University of Lübeck (German):
Project website:
News article BMBF (German):

Project Coordination

Esfandiar Mohammadi, Dr. Photo of Esfandiar  Mohammadi
Institut für IT-Sicherheit (ITS)
+49 451 3101 6609

Szymon Sieciński, Dr. Photo of Szymon Sieciński Sieciński
Institut für Medizinische Informatik
+49 451 3101 5612

Project Members

Abid Hasan, M. Sc. Md Photo of Abid  Hasan
Institut für Medizinische Informatik (IMI)

Laura Liebenow, M. Sc. Photo of Laura  Liebenow
Institut für Medizinische Informatik (IMI)

Jan Wichelmann, M. Sc. Photo of Jan  Wichelmann
Institut für IT-Sicherheit (ITS)
+49 451 3101 6606