Artificial intelligence-informed planning for the rapid response of hazard-impacted road networks.

Post-hazard rapid response has emerged as a promising pathway towards resilient critical infrastructure systems (CISs). Nevertheless, it is challenging to scheme the optimal plan for those rapid responses, given the enormous search space and the hardship of assessment on the spatiotemporal status of CISs. We now present a new approach to post-shock rapid responses of road networks (RNs), based upon lookahead searches supported by machine learning. Following this approach, we examined the resilience-oriented rapid response of a real-world RN across Luchon, France, under destructive earthquake scenarios. Our results show that the introduction of one-step lookahead searches can effectively offset the lack of adaptivity due to the deficient heuristic of rapid responses. Furthermore, the performance of rapid responses following such a strategy is far surpassed, when a series of deep neural networks trained based solely on machine learning, without human interventions, are employed to replace the heuristic and guide the searches.

Urban seismic resilience mapping: a transportation network in Istanbul, Turkey.

When a seismic event occurs, transportation networks play a critical role in undertaking emergency activities such as evacuation and relief supply. Accordingly, to secure their functionality, it is essential to accurately assess their resilience. In particular, this study performs a rigorous probabilistic analysis on the seismic resilience of a transportation network in Istanbul, Turkey. The analysis accuracy is enhanced by considering, along with the structural damage of roadways, the additional disruption mode of network performance caused by the debris falling from damaged objects in their vicinity. Moreover, we obtain the results as a map of resilience measure, which enables us to investigate the disruption inequality across the study area and identify critical factors that govern the system resilience. To enable such sophisticated probabilistic analysis, a Bayesian network (BN) model is developed that involves various types of information from the hazard process to the performance of structures and systems. Then, the BN is quantified by identifying and compiling a comprehensive list of datasets. Thereby, this study analyses large-scale systems involving thousands of structures, while providing general probabilistic models and data schema that can be employed for other transportation networks.

Normal background levels of air and surface mould reserve in English residential building stock: a preliminary study towards benchmarks based on NAHA measurements.

This paper reports results obtained from a surface (both visually clean and dirty/dusty surfaces) and active (aggressive or activated) air testing scheme on 140 residential rooms in England, without visible water damage or mould growth, along with a few rooms with visible mould growth/water damage tested for comparison purposes. The aim was to establish normal background levels of mould in non-water-damaged interiors to benchmark a 'normal' indoor environment, and in turn when there is a need for further investigation, and, possibly, remediation. Air and surface mould was quantified based on the activity of ß-N-acetylhexosaminidase (EC 3.2.1.52; NAHA). The obtained readings showed a log-normal distribution. Ninety-eight percent of the samples obtained from visually clean surfaces were equal to or less than 25 relative fluorescence units (RFU), which is suggested to be the higher bound for the range which can be used as a success criterion for surface cleaning/remediation. Of samples obtained from visually dirty/dusty surfaces, around 98% were below 450 RFU, which is suggested to define the lower-bound for abnormally high levels of mould, rare even on dirty/dusty surfaces. Similarly, around 98% of the air samples were found to have 1700 RFU or below. Values above 1700 RFU are therefore deemed unlikely in a non-problem indoor environment and can be indicative of a possible problem inducing mould growth. The samples with values below 1700 were further divided into three proposed sub-categories. Finally, the obtained RFU values and the suggested benchmarks were compared to those obtained from 17 non-residential indoor environments tested previously in Copenhagen, and the benchmarks that are currently used in Danish national standards, and they were both found to be highly congruent, suggesting that local climate regimes and room functions might not be as influential on indoor mould levels as commonly thought, or that the nuances between England and Denmark in terms of these factors are not strong enough to lead to sizable changes in the typical indoor mould levels in these countries' building stocks.

Surface and passive/active air mould sampling: A testing exercise in a North London housing estate.

Despite indoor mould being one of the most common problems in residential properties in the UK, there are not any widely accepted methodologies for its measurement. This paper focusses on this problem of measurement and reports on the findings from a rigorous testing scheme carried out to quantify air and surface mould concentrations and particle counts within 71 rooms from 64 properties in North London, some with and some without visible mould. The aim was to investigate the potential of passive and active air sampling strategies (sampling from still and actively mixed air, respectively) to explain visible mould, and understand how home/room characteristics correlate with the obtained readings. Airborne mould levels were quantified using an Andersen sampler (passively and actively), as well as by a chemical method based on the quantification of the N-acetylhexosaminidase (NAHA) activity (actively), which was also used to quantify surface mould. The mould levels were then correlated against physical characteristics of the tested homes/rooms, collected by means of survey sheets developed as part of this study. The findings did not reveal any independent variable governing all or most of the response variables, but a complex analysis suggested that whether it is a house or a flat could depict mould levels in the air and on the surfaces. It was also shown that a robust testing protocol should combine air and surface based methods, and an active air sampling strategy leads to a more accurate appraisal of airborne mould levels. Finally, the results showed that while there is some correlation between visible mould (and other moisture induced problems such as condensation) and measured air mould concentrations, lack of visible mould within a room does not necessarily mean low air mould concentrations, and thus one should not rely solely on visual inspection.

Indoor mould testing in a historic building: Blickling Hall.

Indoor mould growth is a growing concern for all stakeholders of built environment, including residents, builders, insurance and building remediation industry as well as custodians of heritage buildings. The National Trust has reported this problem in a number of buildings under their ownership, and developed solutions and fine-tuned their maintenance programme so as to minimise indoor and surface mould growth risk. This paper reports findings from an extensive mould-testing scheme in Blickling Hall, a National Trust property in Norfolk, England, for an appraisal of airborne and surface mould levels within a total of eight rooms, including the famous Long Gallery. The testing protocol used combines active (aggressive) air sampling and surface sampling, analysis of the ß-N-acetylhexosaminidase (NAHA) activity to quantify mould levels and particle counting. The results show that the airborne mould levels are quite low in all spaces, due to satisfactory maintenance of indoor hygrothermal conditions by conservation heating. On the other hand, while the National Trust's developed solutions and maintenance programme have proved effective to avoid surface mould growth in those locations that historically suffered from microbial activity (such as behind book presses, picture frames and tapestries), the results show that the surface cleaning around windows should be improved to tackle surface water due to condensation, which is considered to be the main driver behind high surface NAHA activity obtained in these areas.