In public buildings located in geographies with extreme and highly variable environmental conditions, such as Sweden, managing lighting systems reliably presents unique challenges. These arise not only from technical concerns like lifespan and energy use, but also from the need to ensure human well-being under seasonal extremes of natural light availability. This paper proposes a novel, context-driven framework for asset diagnostics and performance monitoring using two metrics: Critical Integrative Levels (CIL) — which classify lighting zones based on user activity, exposure time, and age group — and Mean Time of Exposure (MTOE), a newly introduced variable that quantifies user interaction with light in space and time. Three case studies across Swedish public libraries (from Luleå to southern Sweden) were used to validate the approach. The results show that integrating human-centric parameters into lighting diagnostics enables more responsive, context-aware threshold setting, compared to traditional functionality-based criteria (e.g., luminous flux percentage). Practical implications include earlier anomaly detection, prioritization of maintenance, and the ability to tailor lighting reliability metrics to real user needs, particularly under non-stationary environmental conditions. The proposed model provides a scalable structure for use in smart public infrastructure, aligning with Industry 5.0 principles. Future research will focus on extending the framework to automated digital twin environments, as well as exploring generalization across sectors such as healthcare, education, and transportation hubs.