emergenCITY researchers investigated whether automatic responses from a smart home are perceived as annoying or helpful in the event of danger.
Leaking gas, intense heat or a power cut: in which dangerous situations would people rely on their smart home’s warning system? Scientists at the LOEWE center emergenCITY asked themselves this question. The results of their studies show that the greater the danger, the more willing people are to let the smart home decide on the right response, such as automatically closing windows and doors. On March 4, the results were published in the journal “Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies”.
The smart home of the future offers comfort and convenience. It automatically heats the rooms in good time before you come home, it lowers the shutters when it is sunny or orders groceries when the fridge is empty. But it can also increase the safety of its residents. At the LOEWE Center emergenCITY, an interdisciplinary research cooperation between the Technical University of Darmstadt, the University of Kassel and the Philipps University of Marburg, scientists are investigating which contribution smart homes can make to warning the population in crises and disasters and responding to dangers.
“While individual devices such as smoke alarms warn residents of certain dangers, integration into comprehensive smart home warning systems offers the potential for greater security, as components can respond to threats based on predefined protocols,” Markus Henkel explains, scientist at emergenCITY and at Science and Technology for Peace and Security PEASEC at TU Darmstadt. In addition, one advantage of a smart home is that it can warn of local dangers as well as passing on official warnings from the authorities.
Together with Darmstadt researchers Steffen Haesler, Hiba Al-Najmi, Frank Hessel and Christian Reuter, he was able to show that people are willing to trust a smart home, especially when there is a threat of great danger. “Then they accept higher levels of automation than in everyday life,” says Markus Henkel. The researchers simulated four different dangerous situations in an energy self-sufficient smart home, the “eHUB”, which is located on the Lichtwiese campus of TU Darmstadt. The 48 participants in the study were given three different options for action, from simple instructions to fully automated execution by the house.
In the highest danger and automation level, for example, they first heard a siren alarm from the smart speaker, saw flashing red light bulbs and strips of light on the handrail of a staircase. In addition, a virtual voice sounded from the smart speaker: “A gas leak has been detected in the area. All doors and windows will be closed automatically.” The smart home then automatically closed the windows and doors. At the lowest level, participants were only told to close windows and doors.
Other hazards included intense heat in the room, in which the house let in fresh air from outside, a simulated power cut, in which it continued to supply electricity thanks to photovoltaics, but also switched off unnecessary power consumers such as ceiling lights or fans, and dry potting soil, in which it watered the flowers itself. The highest level of automation met with the highest level of approval for the greatest danger. For less urgent dangers, the participants wanted to decide for themselves what to do.
In a follow-up study, the researchers found that incorrect reactions from the house influence the decision to completely trust smart home warning systems.
“To increase the acceptance of these systems, they need to be customizable,” Markus Henkel adds.
In order to find out more about the design of such warning systems, the scientists are also investigating in further studies which requirements people with disabilities place on these systems.
Markus Henkel, Steffen Haesler, Hiba K. Al-Najmi, Frank Hessel, and Christian Reuter. 2025. The House That Saves Me? Assessing the Role of Smart Home Automation in Warning Scenarios. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. 9, 1, Article 5 (March 2025), 32 pages. https://doi.org/10.1145/3712269
Mission eHUB: https://www.emergencity.de/missions/ehub/
