Technological innovation has significantly transformed the field of statistics, not only in data analysis but also in data collection. Traditional methods based on direct observation have evolved into hybrid approaches that combine passively collected data (e.g., from GPS or accelerometers) with active user input through digital interfaces. This evolution has led to Smart Surveys—next-generation tools that leverage smart devices, such as smartphones and wearables, to collect data actively (via questionnaires or images) and passively (via embedded sensors). Smart Surveys offer strategic value in industrial contexts by enabling real-time data collection on worker behavior, environments, and operational conditions. However, the heterogeneity of such data poses challenges in management, integration, and quality assurance. This study proposes a modular system architecture incorporating gamification elements to enhance user participation, particularly among hard-to-reach worker segments, such as mobile or shift workers. By leveraging motivational strategies and interactive feedback mechanisms, the system seeks to foster greater engagement while addressing critical data security and privacy concerns within industrial Internet of Things (IoT) environments.

The long-run sustainability of pay-as-you-go pension systems crucially depends on the dynamic balance between social-security contributions paid by the working population and benefits paid to retirees. In Italy, the National Social Security Institute (INPS) manages the core of the public system, whose financial equilibrium is increasingly challenged by demographic aging, labor market fragility, and macroeconomic shocks. In this paper, in line with the aims of the Special Issue “Signal Processing and Machine Learning in Real-Life Processes”, we reinterpret the Italian pension system as a complex stochastic signal-processing problem. Using the most recent data published in the Annuario Statistico Italiano 2024 highlighting by ISTAT—with a focus on Protection and Social Security—we construct a set of time series describing contributions, benefits, coverage ratios and pension amounts, both at the national and territorial level. On this basis, we compare classical time-series models and a recurrent neural network with Long Short-Term Memory (LSTM) architecture for multi-step forecasting of the main aggregates. The signal-processing perspective allows us to disentangle trend, cyclical and shock components, while machine learning provides flexible nonlinear forecasting tools capable of capturing structural breaks such as the COVID-19 crisis. Our empirical results suggest that (i) pension expenditure remains high and persistent as a share of GDP; (ii) the contribution coverage ratio improved in 2022 but remains below the pre-pandemic level; and (iii) regional heterogeneity in the per-capita pension deficit is substantial and stable over time, with persistent imbalances in Southern regions and Islands. Finally, we perform a scenario analysis combining LSTM-based forecasts with demographic and labor market hypotheses, and we quantify the impact of alternative policy measures on the future pension deficit signal. The proposed framework, which integrates permutation-based inference, signal decomposition and deep learning, provides a reproducible template for the real-time monitoring of pension sustainability using official open data.

E-commerce is a key sector in the Italian economy, with online companies becoming some of the largest and most profitable businesses. However, this growth comes with increased risk exposure. This study aims to investigate the relationship between alternative data (contextual factors, Text-Driven Data Enrichment) and the probability of default for Italian e-commerce companies. To date, no studies have examined how these alternative data affect the default probability within this sector. To address this gap, the ongoing research analyzes a dataset of Italian companies, focusing on sector-specific indicators. In the Italian e-commerce market, companies are identified by a unique code that indicates the sales channel but not the types of products marketed. To overcome this limitation, a natural language processing (NLP) model is applied to the es of these companies, allowing us to identify the types of goods or services sold by each company. These enriched data provide a more comprehensive understanding and serve as the foundation for a classification model that uses standard interpretable algorithms for default prediction. Additionally, network structures identified by companies’ similarities are used to extract new insights, which can support decision-making processes for stakeholders at different levels of the supply chain. The model’s performance is evaluated in various scenarios, comparing results with and without the inclusion of alternative data. Key performance metrics are analyzed to demonstrate how integrating alternative data enhances default prediction models.