Attaining and keeping accuracy levels in deployed low cost air quality multisensory systems (LCAQMS) is a challenging and costly task. Current methodologies, in particular, rely on repeated calibrations by colocation with references grade analyzers or, whenever possible, remote imputations. These are respectively affected by logistic costs and deployment delays or high quality remote data availability dependance. This work introduces a novel methodological approach which aims to obtain a continuous, in-place, innetwork recalibration procedure able to attain sufficiently accurate measurements from day zero in high density deployment scenarios while greatly reducing the total costs of ownership and operation. Based on the Federated Learning concept (FL), a limited subset of devices are collocated with reference stations obtaining data capable to produce local calibration models which contribute by harmonization to the development and deployment of a universal calibration model which can be exploited by all the deployed devices, notwithstanding their number. Preliminary, while encouraging, results, based on publicly available dataset exploitation, are here presented.

By imitating the relevant principles of the biological olfactory epithelium layer and the olfactory bulb, a bionic spike coding and a bionic olfactory bulb model are proposed to process the signals of electronic noses. Bionic spike coding converts the sensor's response into spike signals through a process of accumulation and release. This spike signal has similar characteristics to the spikes generated by cells in the biological olfactory epithelium layer, which is more in line with biological systems. The bionic olfactory bulb model is built according to the key cell types and connection pathways found in the biological olfactory bulb. It can effectively simulate the lateral inhibition mechanism of the biological olfactory bulb and achieve contrast enhancement between signals. In order to compare the performance of the proposed method and traditional methods, experiments were first conducted on two collected data sets. When compared with traditional methods, the average classification accuracies of the proposed method reached 95.5 % and 97.2 % respectively, exceeding the sub-optimal methods by 2.1 % and 0.6 % respectively. Experiments were also conducted on three challenging public datasets with the following characteristics: low concentration, small sample size, and highly similar mixed gases. The experiments prove that the proposed method has stronger recognition ability and universality than traditional methods. In addition, other coding methods and bionic olfactory bulb models are also compared. The proposed coding method and bionic olfactory bulb model both achieve the best results, improving by 8.3 % and 1 % respectively compared to the sub-optimal method. (c) 2025 Elsevier Science. All rights reserved.

The increasing adoption of hydrogen as an energy carrier requires advanced monitoring solutions for transport infrastructures, where intelligent sensing and data-driven analysis can play a key role in improving safety and operational efficiency. However, anomaly detection in hydrogen transport networks remains challenging due to the limited availability of operational data and the complexity of transient behaviors associated with these systems. This work investigates a weakly-supervised anomaly detection framework for hydrogen transport networks based on high-fidelity simulation and data-driven analysis. The proposed methodology combines temporal deep learning architectures and unsupervised representation learning models with an operational threshold calibration strategy based on the trade-off between false positives and false negatives. The proposed framework is validated using a high-fidelity simulation environment that reproduces normal and anomalous operating conditions, including leaks, compressor malfunctions, and delayed activation events. The framework is evaluated through comparative experiments involving different anomaly detection architectures, robustness analysis under measurement noise, and leave-one-topology-out generalization tests. Results demonstrate that the proposed approach can effectively identify abnormal behaviors while maintaining robustness against degraded signal quality and previously unseen operating configurations. The obtained results highlight the effectiveness of the proposed methodology as a framework for developing and validating intelligent monitoring strategies for hydrogen transport infrastructures.

Hydrogen transport involves the safe movement of gaseous hydrogen through industrial pipeline networks, typically between production plants, storage facilities, and distribution centers, and is a key component in the transition toward more sustainable energy sources [1]. Monitoring these networks is essential, as hydrogen is highly flammable and leaks, compressor failures, or delayed component responses can lead to serious accidents, environmental damage, and operational interruptions. Despite the growing interest in this sector, publicly available datasets containing multivariate data on hydrogen transport networks are extremely limited, hindering the development and evaluation of data-driven monitoring methods [[2], [3], [4]]. To address this gap, we present a synthetic dataset simulated using a MATLAB Simscape model of a pipeline segment representative of an industrial network [[5], [6], [7],14]. The dataset includes time-series data from distributed virtual sensors, covering both normal operating conditions and anomalous scenarios such as leaks, compressor failures, and delayed component responses [8,9]. The simulation reproduces transient and steady-state dynamics typical of industrial networks, providing data suitable for the development and evaluation of algorithms for digital twins [10], monitoring, and anomaly detection in hydrogen transport infrastructures [10,11].

Outdoor exposure to particulate matter (PM2.5 and PM10) in urban areas can vary considerably depending on the mode of transport. This study aims to quantify this difference in exposure during daily travel, by carrying out a micro-sensor measurement campaign. The pollutant exposure was assessed simultaneously over predefined routes in order to allow comparison between different transport modes having the same starting and ending points. During the six-week measurement campaign, the average reference values for PM background concentrations were 13.72 and 17.92μg/m3 for the PM2.5 and PM10, respectively. The results revealed that the mode with the highest exposure to PM2.5 adjusted to background concentration (PM2.5Norm) was the bus (1.65) followed by metro (1.51), walking (1.33), tramway (1.31), car (1.09) and finally the bike (1.06). For PM10Norm, the tramway had the highest exposure (1.86), followed by walking (1.68), metro (1.65), bus (1.61), bike (1.43) and finally the car (1.39). The level of urbanization around the route and the presence of preferential lanes for public transportation influenced the concentration to which commuters were exposed. For the active modes (bike and walking), we observed frequent variations in concentrations during the trip, characterized by punctual peaks in concentration, depending on the local characteristics of road traffic and urban morphology. Fluctuations in particulate matter inside public transport vehicles were partly explained by the opening and closing of doors during stops, as well as the passenger flows, influencing the re-suspension of particles. The car was one of the least exposed modes overall, with the lowest concentration variability, although these concentrations can vary greatly depending on the ventilation parameters used. These results encourage measures to move the most exposed users away from road traffic, by developing a network of lanes entirely dedicated to cycling and walking, particularly in densely populated areas, as well as encouraging the renewal of motorized vehicles to use less polluting fuels with efficient ventilation systems.

The energy transition relies on an increasingly massive and pervasive use of renewable energy sources, mainly photovoltaic and wind, characterized by an intrinsic degree of production uncertainty, mostly due to meteorological conditions variability that, even if accurately estimated, can hardly be kept under control. Because of this limit, continuously monitoring the operative status of each renewable energy-based power plant becomes relevant in order to timely face any other uncertainty source such as those related to the plant operation and maintenance (O&M), whose effect may become relevant in terms of the levelized cost of energy. In this frame, the use of robots, which incorporate fully automatic platforms capable of monitoring each plant and also allow effective and efficient process operation, can be considered a feasible solution. This paper carries out a review on the use of robots for the O&M of photovoltaic, wind, hydroelectric, and concentrated solar power, including robot applications for controlling power lines, whose role can in fact be considered a key complementary issue within the energy transition. It is shown that various robotic solutions have so far been proposed both by the academy and by industries and that implementing their use should be considered mandatory for the energy transition scenario.

Future air quality monitoring networks will integrate fleets of low-cost gas and particulate matter sensors that are calibrated using machine learning techniques. Unfortunately, it is well known that concept drift is one of the primary causes of data quality loss in machine learning application operational scenarios. The present study focuses on addressing the calibration model update of low-cost NO2 sensors once they are triggered by a concept drift detector. It also defines which data are the most appropriate to use in the model updating process to gain compliance with the relative expanded uncertainty (REU) limits established by the European Directive. As the examined methodologies, the general/global and the importance weighting calibration models were applied for concept drift effects mitigation. Overall, for all the devices under test, the experimental results show the inadequacy of both models when performed independently. On the other hand, the results from the application of both models through a stacking ensemble strategy were able to extend the temporal validity of the used calibration model by three weeks at least for all the sensor devices under test. Thus, the usefulness of the whole information content gathered throughout the original co-location process was maximized.

Scalable and effective calibration is a fundamental requirement for low-cost air quality (AQ) monitoring systems and will enable accurate and pervasive monitoring in cities. Suffering from environmental interferences and fabrication variance, these devices need to encompass sensor-specific and complex calibration processes for reaching a sufficient accuracy to be deployed as indicative measurement devices in AQ monitoring networks. Concept and sensor drift often force the calibration process to be frequently repeated. These issues lead to unbearable calibration costs, which denies their massive deployment when accuracy is a concern. In this work, we propose a zero transfer samples, global calibration methodology as a technological enabler for Internet of Things (IoT) AQ multisensory devices, which relies on low-cost particulate matter (PM) sensors. This methodology is based on field recorded responses from a limited number of IoT AQ multisensors units and machine learning (ML) concepts and can be universally applied to all units of the same type. A multiseason test campaign has shown that, when applied to different sensors, this methodology's performances match those of state-of-the-art methodology, which requires to derive different calibration parameters for each different unit. If confirmed, these results show that, when properly derived, a global calibration law can be exploited for a large number of networked devices with dramatic cost reduction eventually allowing massive deployment of accurate IoT AQ monitoring devices. Furthermore, this calibration model could be easily embedded onboard of the device or implemented on the edge allowing immediate access to accurate readings for personal exposure monitor applications as well as reducing long-range data transfer needs.

The last decade has seen a significant growth in the adoption of low-cost air quality monitoring systems (LCAQMSs), mostly driven by the need to overcome the spatial density limitations of traditional regulatory grade networks. However, urban air quality monitoring scenarios have proved extremely challenging for their operative deployment. In fact, these scenarios need pervasive, accurate, personalized monitoring solutions along with powerful data management technologies and targeted communications tools; otherwise, these scenarios can lead to a lack of stakeholder trust, awareness, and, consequently, environmental inequalities. The AirHeritage project, funded by the EU’s Urban Innovative Action (UIA) program, addressed these issues by integrating intelligent LCAQMSs with conventional monitoring systems and engaging the local community in multi-year measurement strategies. Its implementation allowed us to explore the benefits and limitations of citizen science approaches, the logistic and functional impacts of IoT infrastructures and calibration methodologies, and the integration of AI and geostatistical sensor fusion algorithms for mobile and opportunistic air quality measurements and reporting. Similar research or operative projects have been implemented in the recent past, often focusing on a limited set of the involved challenges. Unfortunately, detailed reports as well as recorded and/or cured data are often not publicly available, thus limiting the development of the field. This work openly reports on the lessons learned and experiences from the AirHeritage project, including device accuracy variance, field recording assessments, and high-resolution mapping outcomes, aiming to guide future implementations in similar contexts and support repeatability as well as further research by delivering an open datalake. By sharing these insights along with the gathered datalake, we aim to inform stakeholders, including researchers, citizens, public authorities, and agencies, about effective strategies for deploying and utilizing LCAQMSs to enhance air quality monitoring and public awareness on this challenging urban environment issue.

Photovoltaic energy (PV) is considered one of the pillars of the energy transition. However, this energy source is limited by a power density per unit surface lower than 200 W/m2, depending on the latitude of the installation site. Compared to fossil fuels, such low power density opens a sustainability issue for this type of renewable energy in terms of its competition with other land uses, and forces us to consider areas suitable for the installation of photovoltaic arrays other than farmlands. In this frame, floating PV plants, installed in internal water basins or even offshore, are receiving increasing interest. On the other hand, this kind of installation might significantly affect the water ecosystem environment in various ways, such as by the effects of solar shading or of anchorage installation. As a result, monitoring of floating PV (FPV) plants, both during the ex ante site evaluation phase and during the operation of the PV plant itself, is therefore necessary to keep such effects under control. This review aims to examine the technical and academic literature on FPV plant monitoring, focusing on the measurement and discussion of key physico-chemical parameters. This paper also aims to identify the additional monitoring features required for energy assessment of a floating PV system compared to a ground-based PV system. Moreover, due to the intrinsic difficulty in the maintenance operations of PV structures not installed on land, novel approaches have introduced autonomous solutions for monitoring the environmental impacts of FPV systems. Technologies for autonomous mapping and monitoring of water bodies are reviewed and discussed. The extensive technical literature analyzed in this review highlights the current lack of a cohesive framework for monitoring these impacts. This paper concludes that there is a need to establish general guidelines and criteria for standardized water quality monitoring (WQM) and management in relation to FPV systems.

The accurate prediction of photovoltaic (PV) energy production is a crucial task to optimise the integration of solar energy into the power grid and maximise the benefit of renewable source trading in the energy market. This paper systematically and quantitatively analyses the literature by comparing different machine learning techniques and the impact of different meteorological forecast providers. The methodology consists of an irradiance model coupled with a meteorological provider; this combination removes the constraint of a local irradiance measurement. The result is a Transformer Neural Network architecture, trained and tested using OpenMeteo data, whose performance is superior to other combinations, providing a MAE of 1.22 kW (0.95%), and a MAPE of 2.21%. The implications of our study suggest that adopting a comprehensive approach, integrating local weather data, modelled irradiance, and PV plant configuration data, can significantly improve the accuracy of PV power forecasting, thus contributing to more effective technological and economic integration.

There is a consensus within the scientific community regarding the effects on the environment, health, and climate of the use of renewable energy sources, which is characterized by a rate of harmful polluting emissions that is significantly lower than that typical of fossil fuels. On the other hand, this transition towards the use of more sustainable energy sources will also be characterized by an increasingly widespread electrification rate. In this work, we want to discuss whether electricity distribution and transmission networks and their main components are characterized by emissions that are potentially harmful to the environment and human health during their operational life. We will see that the scientific literature on this issue is rather limited, at least until now. However, conditions are reported in which the network directly causes or at least promotes the emissions of polluting substances into the environment. For the most part, the emissions recorded, rather than their environmental or human health impacts, are studied as part of the implementation of techniques for the early determination of faults in the network. It is probable that with the increasing electrification of energy consumption, the problem reported here will become increasingly relevant.

Recently, the state of the art field calibration approach for low cost air quality sensors have been criticized fort its lack of scalability. Actually the need for co-locating with reference analyzers each and every multi-sensors for a significant amount of time is definitely unfeasible if a truly pervasive commercial deployment of this technology is pursued. Remote calibration using post-deployment data stream along with those coming from nearby reference stations open source data can be a viable solution. This methodology can actually cope with sensors accuracy issues and can limit the effects of concept and sensors drift by continuous iteration of the calibration procedure. However, the imperfect match of the data recorded in different places which is inherent of the air quality spatial variance poses serious doubt on the generalization of this procedure. Careful reference station selection and data management is needed to extract the beneficial informative content from its open source data stream. Using a multi-seasonal multi-unit muti-location dataset, this work compares the performance obtained by different remote calibration strategies comparing them with state-of-the-art approaches quantifying the benefit obtainable by such methodologies and the performance impact of the calibration design choices.

There is an increasing scientific interest in studying vehicular traffic pollution in road tunnels. This is due both to the interest in evaluating the effect that the different polluting gases can have on the driving style of motorists and also to the hypothesis that tunnels could be considered as closed systems in which the vehicular traffic–pollution correlation is easier to study because it is more easily separated from other effects. In this work, a system of low-cost IoT sensor nodes for the detection of carbon monoxide (CO), nitrogen dioxide (NO2), ozone (O3), particulate matters (PM1, PM2.5, PM10), relative humidity (RH) and temperature (T) has been installed in an Italian tunnel, where vehicular traffic has been measured and classified for type of vehicles. The results of the measurement campaign, which lasted 3 months, from April to June 2022, allowed us to state that road tunnels actually behave like closed and isolated systems in which pollution may be directly correlated to the traffic volume and type. Furthermore, data show that quite high values of the major pollutants are observable in the tunnel in comparison to the external environment. As such, IoT sensor nodes may contribute to a distributed measuring approach on the road tunnel system mechanics assessment including, as an example, the operational impacts of forced ventilation.

Air Quality is of significant concern in modern life. Since pollutants are harmful for health and environment, it is important to measure their concentration in the air. In this work the authors report on a tool for the measurement of the concentration of CO, NO2, O3, VOC and PM in the air in harsh environments such as motorway tunnels or underground transportation system. Inside a tunnel, pollution is for the major part composed of gasses released by vehicles and particulate matters, mainly due to vehicle’s brake particles, asphalt erosion and tires wearing. The tool is composed of a network of rugged low-cost solid-state sensor nodes placed in the environment to continuously monitor the gas concentrations and a remote backend that stores, handles and presents the information received via cellular network. The sensor network has been installed in a roadway tunnel in southern Sicily (Italy) and here the authors report preliminary results related to the continuous measurement of CO and NO2. The aim of this work is to contribute to make continuous the monitoring of gases concentrations inside a roadway tunnel. The sensors have been designed to be online for a long period, allowing remote maintenance. The results will be deeply investigated after a long period of data collection. The analysis of data will be useful to pose in place appropriate actions to mitigate dangerous concentrations of gases.

Particulate matter (PM) in air has been proven to be hazardous to human health. Here we focused on analysis of PM data we obtained from the same campaign which was presented in our previous study. Multivariate linear and random forest models were used for the calibration and analysis. In our linear regression model the inputs were PM, temperature and humidity measured with low-cost sensors, and the target was the reference PM measurements obtained from SEPA in the same timeframe.

Recent advances in IoT and chemical sensors calibration technologies have led to the proposal of Hierarchical air quality monitoring networks. They are indeed complex systems relying on sensing nodes which differs from size, cost, accuracy, technology, maintenance needs while having the potential to empower smart cities and communiities with increased knowledge on the highly spatiotemporal variance Air Quality phenomenon (see [1]). The AirHeritage project, funded by Urban Innovative Action program have developed and implemented a hierarchical monitoring system which allows for offering real time assessments and model based forecasting services including 7 fixed low cost sensors station, one (mobile and temporary located) regulatory grade analyzer and a citizen science based ultra high resolution AQ mapping tool based on field calibrated mobile analyzers. This work will analyze the preliminary results of the project by focusing on the machine learning driven sensors calibration methodology and citizen science based air quality mapping campaigns. Thirty chemical and particulate matter multisensory devices have been deployed in Portici, a 4Km2 city located 7 km south of Naples which is affected by significant car traffic. The devices have been entrusted to local citizens association for implementing 1 preliminary validation campaign (see [2]) and 3 opportunistic 2-months duration monitoring campaigns. Each 6 months, the devices undergoes a minimum 3 weeks colocation period with a regulatory grade analyzer allowing for training and validation dataset building. Multilinear regression sw components are trained to reach ppb level accuracy (MAE <10ug/m^3 for NO2 and O3, <15ug/M^3 for PM2.5 and PM10, <300ug/M^3 for CO) and encoded in a companion smartphone APP which allows the users for real time assessment of personal exposure. In particular, a novel AQI strongly based on European Air Quality Index ([3]) have been developed for AQ real time data communication. Data have been collected using a custom IoT device management platform entrusted with inception, storage and data-viz roles. Finally data have been used to build UHR (UHR) AQ maps, using spatial binning approach (25mx25m) and median computation for each bin receiving more than 30 measurements during the campaign. The resulting maps have hown the possibility to allow for pinpointing city AQ hotpots which will allows fact-based remediation policies in cities lacking objective technologies to locally assess concentration exposure. [1] Nuria Castell et Al., Can commercial low-cost sensor platforms contribute to air quality monitoring and exposure estimates?, Environment International, Volume 99, 2017, Pages 293-302 ISSN 0160-4120, https://doi.org/10.1016/j.envint.2016.12.007. [2] De Vito, S, et al., Crowdsensing IoT Architecture for Pervasive Air Quality and Exposome Monitoring: Design, Development, Calibration, and Long-Term Validation. Sensors 2021, 21, 5219. https://doi.org/10.3390/s21155219 [3] https://airindex.eea.europa.eu/Map/AQI/Viewer/

Field calibration is recognized as the best performing approach for operational deployment of low cost air quality sensors. In this approach, each individual sensors node is co-located with reference analyzers to derive ad-hoc calibration algorithms parameters. This strongly limits scalability, and thus mass deployment of these devices for improved air quality phenomena knowledge and management. Global calibration, aiming to derive a single calibration function to be used for all multisensors in a batch, is a promising approach recently tackled in artificial olfaction field. In this work, this approach is firstly applied to low cost PM sensors. Results show comparable performances with conventional, ad-hoc (individual) approaches paving the way for a more scalable approach to calibration of AQ multisensors devices.