Satakunta University of Applied Sciences

Smart spaces are ubiquitous computing environments that integrate diverse sensing and communication technologies to enhance space functionality, optimize energy utilization, and improve user comfort and well-being. The integration of emerging AI methodologies into these environments facilitates the formation of AI-driven smart spaces, which further enhance functionalities of the spaces by enabling advanced applications such as personalized comfort settings, interactive living spaces, and automatization of the space systems, all resulting in enhanced indoor experiences of the users. In this paper, we present a systematic survey of existing research on the foundational components of AI-driven smart spaces, including sensor technologies, data communication protocols, sensor network management and maintenance strategies, as well as the data collection, processing and analytics. Given the pivotal role of AI in establishing AI-powered smart spaces, we explore the opportunities and challenges associated with traditional machine learning (ML) approaches, such as deep learning (DL), and emerging methodologies including large language models (LLMs). Finally, we provide key insights necessary for the development of AI-driven smart spaces, propose future research directions, and sheds light on the path forward.

Infectious diseases continue to impose substantial health and economic burdens on societies, and the built environment has become an increasingly important target for infection prevention and control. In response, a growing range of human-independent indoor hygiene solutions has been developed, including antimicrobial surfaces and coatings, touchless fixtures, air cleaning devices, and antimicrobial lighting. These interventions are attractive because, unlike conventional hygiene measures, they do not rely primarily on user compliance. However, the evidence base for these technologies remains uneven. Most studies focus on reductions in microbial contamination, whereas far fewer assess outcomes that are more meaningful for decision-making, such as infection incidence, absenteeism, or healthcare utilization. From a public health and built-environment perspective, this distinction is crucial: a lower microbial burden does not necessarily translate into reduced morbidity. In this Perspective, we propose a structured framework for evaluating human-independent indoor hygiene solutions that integrates financial, health-related, and environmental costs alongside potential benefits. Assessments should consider not only health gains but also financial costs, unintended health effects, environmental burdens, and opportunity costs. A nursing home pilot is used as an illustrative example of how these cost categories emerge in practice. Current evidence suggests that consistent morbidity reduction has not yet been demonstrated for most human-independent indoor hygiene solutions. We therefore argue that their evaluation and implementation should move beyond proxy microbiological indicators and adopt a structured, context-sensitive approach. Until stronger evidence is available, their implementation should remain cautious and selective.

IntroductionEffective infection prevention and control (IPC) in buildings requires integrated strategies addressing air, surfaces, and water systems.MethodsThis study evaluated the impact of an Indoor Hygiene Concept (IHC) on microbial contamination and infection-related absenteeism in a real-world kindergarten setting. A Living Lab intervention was conducted, where one unit implemented antimicrobial coatings, antimicrobial blue light technology, and a portable air purifier, while a comparable control unit remained unchanged. Environmental sampling of high-touch surfaces (n = 368 per unit), monitoring of infection-related absences among children aged 3–5 years and staff, and continuous indoor air quality assessment were performed between 2023 and 2025.ResultsBaseline environmental and hygiene conditions were comparable between units. Antimicrobial coatings reduced mean surface microbial load (9.4 vs. 15.9 CFU/cm2), although median values were similar, indicating very limited effects under high baseline hygiene conditions. The intervention unit showed a consistent but not statistically significant reduction in infection-related absenteeism (19% in staff and 12% in children).DiscussionThese results suggest that integrated indoor hygiene solutions may support infection control in real-world settings. More extensive studies, including larger human study populations, and viral measurements are needed to confirm these preliminary observations and evaluate long-term sustainability.

Smart environments increasingly integrate artificial intelligence (AI) to process sensor data, coordinate Internet of Things (IoT) devices, and actuate cyber-physical services.Within these environments, large language models (LLMs) are emerging as decision-making and interaction components. However, LLM integration expands the attack surface of these environments, and existing benchmarks focus largely on risks associated with text-based interactions and LLM-generated outputs, failing to capture cyber-physical risks that arise when models control physical actuators. This paper introduces Evil-AI Benchmark, an open-source evaluation framework for assessing LLM agents in smart environments through (i) five adversarial threat vectors—prompt injection, persuasion, attacker-in-the-middle (AITM) simulation, data leakage, and unsafe actions—and (ii) a capability validation that checks whether a model can correctly activate device-control tools. The benchmark instruments tool activations and forwards tool-trigger events to an external Arduino-driven servo actuator, providing physical verification of actuation events. To demonstrate the benchmark in practice, we evaluate eight representative LLMs using 350 executable scenarios: 250 adversarial tests spanning five smart-environment domains (smart home, healthcare IoT, industrial control, public infrastructure, and smart building), 50 capability checks, and 50 benign over-refusal tests. Responses are scored by two independent judge models and verified by a human evaluator. We report Evilness as the extent to which attacks succeed. The Evilness Score equals the number of successful attacks, and the Evilness Rate is the corresponding percentage. Across models, Evilness Rates range from 0.8% to 54.0% (2–135 successful attacks out of 250), while Defense Rates range from 46.0% to 99.2%. The most persistent weaknesses occur in persuasion, especially under sustained multi-turn pressure, and in prompt injection. Moreover, larger models are not necessarily more robust against adversarial attacks. We additionally quantify over-refusal on benign authorized requests, so that safety is not rewarded by indiscriminate refusal. Evil-AI Benchmark provides reproducible diagnostics and quantitative metrics that connect text-based safety evaluation with real-world cyber-physical operation.

Social inclusion shapes well-being, yet people with communication or functional challenges often face barriers to participation. Smart clothing offers discreet ways to support interaction, but little is known about how broader user groups envision such technologies. This study explored expectations for cloth-based activators through two co-creation workshops with 24 participants. Researcher notes from the sessions were deductively analysed to identify user-driven priorities. Participants proposed several concepts that converged around three themes. First, they envisioned cloth-based activators as subtle tools for communication-enabling quiet emotional cues, discreet requests for help, or simple affirmations. Second, personalisation was essential: users expected adjustable functions, flexible placement, and aesthetic alignment with individual preferences. Third, participants saw these technologies as everyday companions that could support task management, emotional regulation or daily coordination with minimal disruption. Overall, the findings suggest that textile-integrated activators may promote social inclusion not through complex functions, but through simple, embodied and user-driven interactions embedded in clothing. Early-stage co-creation revealed socially grounded expectations that can guide the development of personalised and unobtrusive smart textile technologies.