An End-to-End Artificial Intelligence of Things (AIoT) Solution for Protecting Pipeline Easements against External Interference-An Australian Use-Case.

High-pressure pipelines are critical for transporting hazardous materials over long distances, but they face threats from third-party interference activities. Preventive measures are implemented, but interference accidents can still occur, making the need for high-quality detection strategies vital. This paper proposes an end-to-end Artificial Intelligence of Things (AIoT) solution to detect potential interference threats in real time. The solution involves developing a smart visual sensor capable of processing images using state-of-the-art computer vision algorithms and transmitting alerts to pipeline operators in real time. The system's core is based on the object-detection model (e.g., You Only Look Once version 4 (YOLOv4) and DETR with Improved deNoising anchOr boxes (DINO)), trained on a custom Pipeline Visual Threat Assessment (Pipe-VisTA) dataset. Among the trained models, DINO was able to achieve the best Mean Average Precision (mAP) of 71.2% for the unseen test dataset. However, for the deployment on a limited computational-ability edge computer (i.e., the NVIDIA Jetson Nano), the simpler and TensorRT-optimized YOLOv4 model was used, which achieved a mAP of 61.8% for the test dataset. The developed AIoT device captures the image using a camera, processes on the edge using the trained YOLOv4 model to detect the potential threat, transmits the threat alert to a Fleet Portal via LoRaWAN, and hosts the alert on a dashboard via a satellite network. The device has been fully tested in the field to ensure its functionality prior to deployment for the SEA Gas use-case. The AIoT smart solution has been deployed across the 10km stretch of the SEA Gas pipeline across the Murray Bridge section. In total, 48 AIoT devices and three Fleet Portals are installed to ensure the line-of-sight communication between the devices and portals.

The Role of Hydrogen on the Behavior of Intergranular Cracks in Bicrystalline α-Fe Nanowires.

Hydrogen embrittlement (HE) has been extensively studied in bulk materials. However, little is known about the role of H on the plastic deformation and fracture mechanisms of nanoscale materials such as nanowires. In this study, molecular dynamics simulations are employed to study the influence of H segregation on the behavior of intergranular cracks in bicrystalline α-Fe nanowires. The results demonstrate that segregated H atoms have weak embrittling effects on the predicted ductile cracks along the GBs, but favor the cleavage process of intergranular cracks in the theoretically brittle directions. Furthermore, it is revealed that cyclic loading can promote the H accumulation into the GB region ahead of the crack tip and overcome crack trapping, thus inducing a ductile-to-brittle transformation. This information will deepen our understanding on the experimentally-observed H-assisted brittle cleavage failure and have implications for designing new nanocrystalline materials with high resistance to HE.

Optimisation of dispersion parameters of Gaussian plume model for CO2 dispersion.

The carbon capture and storage (CCS) and enhanced oil recovery (EOR) projects entail the possibility of accidental release of carbon dioxide (CO2) into the atmosphere. To quantify the spread of CO2 following such release, the 'Gaussian' dispersion model is often used to estimate the resulting CO2 concentration levels in the surroundings. The Gaussian model enables quick estimates of the concentration levels. However, the traditionally recommended values of the 'dispersion parameters' in the Gaussian model may not be directly applicable to CO2 dispersion. This paper presents an optimisation technique to obtain the dispersion parameters in order to achieve a quick estimation of CO2 concentration levels in the atmosphere following CO2 blowouts. The optimised dispersion parameters enable the Gaussian model to produce quick estimates of CO2 concentration levels, precluding the necessity to set up and run much more complicated models. Computational fluid dynamics (CFD) models were employed to produce reference CO2 dispersion profiles in various atmospheric stability classes (ASC), different 'source strengths' and degrees of ground roughness. The performance of the CFD models was validated against the 'Kit Fox' field measurements, involving dispersion over a flat horizontal terrain, both with low and high roughness regions. An optimisation model employing a genetic algorithm (GA) to determine the best dispersion parameters in the Gaussian plume model was set up. Optimum values of the dispersion parameters for different ASCs that can be used in the Gaussian plume model for predicting CO2 dispersion were obtained.

Experiment and molecular dynamics simulation of nanoindentation of body centered cubic iron.

Experiments and molecular dynamics (MD) simulations have been conducted to investigate the nanoindentation behaviours of iron with body centered cubic (BCC) structure. The experiments show that the indentation hardness decreases with the indentation depth and it changes sharply for a small depth. Two cases with different crystallographic orientations have been simulated. The indentation plane is (010) for Case I and (111) for Case II, respectively. The calculated harness (17.4 GPa for Case I and 22.6 GPa for Case II) are in reasonable agreement with the experimental value (24.2 GPa). The simulation results show that the crystallographic orientation significantly influences the indentation deformation. Case I and Case II exhibit different deformation patterns. The indentation force and the hardness in Case I are smaller than Case II. It is also found that the pileup around the indenter is mainly formed along [110] direction for both cases.

System modelling of a lateral force microscope.

To quantitatively analyse lateral force microscope measurements one needs to develop a model able to relate the photodiode signal to the force acting on the tip apex. In this paper we focus on the modelling of the interaction between the cantilever and the optical chain. The laser beam is discretized by a set of rays which propagates in the system. The analytical equation of a single ray's position on the optical sensor is presented as a function of the reflection's state on top of the cantilever. We use a finite element analysis on the cantilever to connect the optical model with the force acting on the tip apex. A first-order approximation of the constitutive equations are derived along with a definition of the system's crosstalk. Finally, the model is used to analytically simulate the 'wedge method' in the presence of crosstalk in 2D. The analysis shows how the torsion loop and torsion offset signals are affected by the crosstalk.