Time-critical systems must satisfy temporal constraints whose correctness depends not only on event ordering but also on precise timing. Metric Interval Temporal Logic (MITL) provides a formalism to express such requirements. Although robustness has been widely studied under signal-based interpretations, it remains largely unexplored for point-based semantics, where executions are sequences of timestamped facts. In this setting, small timing variations may arbitrarily change Boolean satisfaction, revealing the instability of temporal truth under uncertainty. We introduce a notion of time robustness for MITL over point-based semantics, interpreting robustness as a margin of validity of temporal interpretations. We define a quantitative semantics and prove soundness with respect to Boolean satisfaction together with a Lipschitz stability property with respect to timestamp perturbations, which induces a metric notion of proximity between interpretations. The semantics admits a polynomial-time evaluation procedure and is illustrated on two case studies (drone surveillance and smart hospital), where robustness empirically correlates with tolerance to temporal noise.

Federated Learning has emerged as a promising paradigm that enables collaborative model training while preserving data privacy, thus contributing to enhance user trust. However, the design of Federated Learning systems requires non-trivial architectural choices to address several challenges, such as system efficiency and learning accuracy. Architectural patterns for Federated Learning systems have been defined in the literature to handle these challenges, but their experimentation is limited. The objective of this paper is to empower software architects in their task of evaluating the design of FL systems while deciding which architectural alternatives are more beneficial in their context of adoption. Our methodology consists of a tool-based approach that embeds the implementation of six architectural patterns defined in the literature. The advantage is that software architects can select design alternatives either in isolation or in a combined fashion, and the subsequent analysis provides the evaluation of some metrics of interest. The experimental results indicate that architectural patterns can enhance system efficiency, although we found a combination of patterns that added overhead and turned to limit its benefit. By quantifying these trade-offs, we aim to support software architects in designing Federated Learning systems by evaluating the benefits and drawbacks of applying architectural patterns.