Search results

This paper provides practical advice for archaeologists and heritage specialists wishing to use ML approaches to identify archaeological features in high-resolution satellite imagery (or other remotely sensed data sources). We seek to balance the disproportionately optimistic literature related to the application of ML to archaeological prospection through a discussion of limitations, challenges and other difficulties. We further seek to raise awareness among researchers of the time, effort, expertise and resources necessary to implement ML successfully, so that they can make an informed choice between ML and manual inspection approaches.

Automated object detection has been the holy grail of archaeological remote sensing for the last two decades. Machine learning (ML) models have proven able to detect uniform features across a consistent background, but more variegated imagery remains a challenge. We set out to detect burial mounds in satellite imagery from a diverse landscape in Central Bulgaria using a pre-trained Convolutional Neural Network (CNN) plus additional but low-touch training to improve performance. Training was accomplished using MOUND/NOT MOUND cutouts, and the model assessed arbitrary tiles of the same size from the image. Results were assessed using field data.

Validation of results against field data showed that self-reported success rates were misleadingly high, and that the model was misidentifying most features. Setting an identification threshold at 60% probability, and noting that we used an approach where the CNN assessed tiles of a fixed size, tile-based false negative rates were 95–96%, false positive rates were 87–95% of tagged tiles, while true positives were only 5–13%. Counterintuitively, the model provided with training data selected for highly visible mounds (rather than all mounds) performed worse. Development of the model, meanwhile, required approximately 135 person-hours of work.

Our attempt to deploy a pre-trained CNN demonstrates the limitations of this approach when it is used to detect varied features of different sizes within a heterogeneous landscape that contains confounding natural and modern features, such as roads, forests and field boundaries. The model has detected incidental features rather than the mounds themselves, making external validation with field data an essential part of CNN workflows. Correcting the model would require refining the training data as well as adopting different approaches to model choice and execution, raising the computational requirements beyond the level of most cultural heritage practitioners.

Improving the pre-trained model’s performance would require considerable time and resources, on top of the time already invested. The degree of manual intervention required – particularly around the subsetting and annotation of training data – is so significant that it raises the question of whether it would be more efficient to identify all of the mounds manually, either through brute-force inspection by experts or by crowdsourcing the analysis to trained – or even untrained – volunteers. Researchers and heritage specialists seeking efficient methods for extracting features from remotely sensed data should weigh the costs and benefits of ML versus manual approaches carefully.

Our literature review indicates that use of artificial intelligence (AI) and ML approaches to archaeological prospection have grown exponentially in the past decade, approaching adoption levels associated with “crossing the chasm” from innovators and early adopters to the majority of researchers. The literature itself, however, is overwhelmingly positive, reflecting some combination of publication bias and a rhetoric of unconditional success. This paper presents the failure of a good-faith attempt to utilise these approaches as a counterbalance and cautionary tale to potential adopters of the technology. Early-majority adopters may find ML difficult to implement effectively in real-life scenarios.

Unlike many high-profile reports from well-funded projects, our paper represents a serious but modestly resourced attempt to apply an ML approach to archaeological remote sensing, using techniques like transfer learning that are promoted as solutions to time and cost problems associated with, e.g. annotating and manipulating training data. While the majority of articles uncritically promote ML, or only discuss how challenges were overcome, our paper investigates how – despite reasonable self-reported scores – the model failed to locate the target features when compared to field data. We also present time, expertise and resourcing requirements, a rarity in ML-for-archaeology publications.

The purpose of this paper is to use the corresponding magnetic sensor and detection method to detect and image the defects of small diameter pipelines. Urban gas pipeline is an energy transportation tool for urban industrial production and social life, which is closely related to urban safety. Preventing the occurrence of urban gas pipeline transportation accidents and carrying out pipeline defect detection are of great significance for the urban economic and social stability. To perform pipeline defect detection, the magnetic flux leakage internal detection method is generally used in the detection of large-diameter long-distance oil and gas pipelines. However, in terms of the internal detection of small-diameter pipelines, due to the heavy weight, large structure of the detection device and small pipe diameter, the detection is more difficult.

In order to solve the above matters, self-made three-dimensional magnetic sensor and three-dimensional magnetic flux leakage imaging direct method are proposed for studying the defect identification. Firstly, for adapting to the diameter range of small-diameter pipelines, and containing the complete information of the defect, a self-made three-dimensional magnetic sensor is made in this paper to improve the accuracy of magnetic flux leakage detection. And on the basis of it, a small diameter pipeline defect detection system is built. Secondly, as detection signal may be affected by background magnetic field interference and the jitter interference, the complete ensemble empirical mode decomposition with adaptive noise method is utilized to screen the detected signal. As a result, the useful signal is reconstructed and the interference signal is removed. Finally, the defect contour inversion imaging of detection is realized based on the direct method of three-dimensional magnetic flux leakage imaging, which includes three-dimensional magnetic flux leakage detection data and data segmentation recognition.

The three-dimensional magnetic flux leakage imaging experimental results shown that, compared to the actual defects, the typical defects, irregular defects and crack groove defects can be analyzed by the magnetic flux leakage defect contour imaging method in qualitative and quantitative way respectively, which provides a new idea for the research of defect recognition.

A three-dimensional magnetic sensor is made to adapt the diameter range of small diameter pipeline, and based on it, a small-diameter pipeline defect detection system is built to collect and display the magnetic flux leakage signal.

This study aims to propose and evaluate the progress in the basic-pixel (a strategy to generate continuous trajectories that fill out the entire surface) algorithm towards performance gain. The objective is also to investigate the operational efficiency and effectiveness of an enhanced version compared with conventional strategies.

For the first objective, the proposed methodology is to apply the improvements proposed in the basic-pixel strategy, test it on three demonstrative parts and statistically evaluate the performance using the distance trajectory criterion. For the second objective, the enhanced-pixel strategy is compared with conventional strategies in terms of trajectory distance, build time and the number of arcs starts and stops (operational efficiency) and targeting the nominal geometry of a part (operational effectiveness).

The results showed that the improvements proposed to the basic-pixel strategy could generate continuous trajectories with shorter distances and comparable building times (operational efficiency). Regarding operational effectiveness, the parts built by the enhanced-pixel strategy presented lower dimensional deviation than the other strategies studied. Therefore, the enhanced-pixel strategy appears to be a good candidate for building more complex printable parts and delivering operational efficiency and effectiveness.

This paper presents an evolution of the basic-pixel strategy (a space-filling strategy) with the introduction of new elements in the algorithm and proves the improvement of the strategy’s performance with this. An interesting comparison is also presented in terms of operational efficiency and effectiveness between the enhanced-pixel strategy and conventional strategies.

This paper aims to propose a stable and precise recognition and pose estimation method to deal with the difficulties that industrial parts often present, such as incomplete point cloud data due to surface reflections, lack of color texture features and limited availability of effective three-dimensional geometric information. These challenges lead to less-than-ideal performance of existing object recognition and pose estimation methods based on two-dimensional images or three-dimensional point cloud features.

In this paper, an image-guided depth map completion method is proposed to improve the algorithm's adaptability to noise and incomplete point cloud scenes. Furthermore, this paper also proposes a pose estimation method based on contour feature matching.

Through experimental testing on real-world and virtual scene dataset, it has been verified that the image-guided depth map completion method exhibits higher accuracy in estimating depth values for depth map hole pixels. The pose estimation method proposed in this paper was applied to conduct pose estimation experiments on various parts. The average recognition accuracy in real-world scenes was 88.17%, whereas in virtual scenes, the average recognition accuracy reached 95%.

The proposed recognition and pose estimation method can stably and precisely deal with the difficulties that industrial parts present and improve the algorithm's adaptability to noise and incomplete point cloud scenes.

Productivity is often cited as a key barrier to the adoption of metal laser-based powder bed fusion (ML-PBF) technology for mass production. Newer generations of this technology work to overcome this by introducing more lasers or dramatically different processing techniques. Current generation ML-PBF machines are typically not capable of taking on additional hardware to maximise productivity due to inherent design limitations. Thus, any increases to be found in this generation of machines need to be implemented through design or adjusting how the machine currently processes the material. The purpose of this paper is to identify the most beneficial existing methodologies for the optimisation of productivity in existing ML-PBF equipment so that current users have a framework upon which they can improve their processes.

The review method used here is the preferred reporting items for systematic review and meta-analysis (PRISMA). This is complemented by using an artificial intelligence-assisted literature review tool known as Elicit. Scopus, WEEE, Web of Science and Semantic Scholar databases were searched for articles using specific keywords and Boolean operators.

The PRIMSA and Elicit processes resulted in 51 papers that met the criteria. Of these, 24 indicated that by using a design of experiment approach, processing parameters could be created that would increase productivity. The other themes identified include scan strategy (11), surface alteration (11), changing of layer heights (17), artificial neural networks (3) and altering of the material (5). Due to the nature of the studies, quantifying the effect of these themes on productivity was not always possible. However, studies citing altering layer heights and processing parameters indicated the greatest quantifiable increase in productivity with values between 10% and 252% cited. The literature, though not always explicit, depicts several avenues for the improvement of productivity for current-generation ML-PBF machines.

This systematic literature review provides trends and themes that aim to influence and support future research directions for maximising the productivity of the ML-PBF machines.

The accuracy and reliability of upper limb motion assessment have received great attention in the field of rehabilitation. Grasping test is widely carried out for motion assessment, which requires patients to grasp objects and move them to target place. The traditional assessments test the upper limb motion ability by therapists, which mainly relies on experience and lacks quantitative indicators. This paper aims to propose a deep learning method based on the vision system of our upper limb rehabilitation robot to recognize the motion trajectory of rehabilitation target objects automatically and quantitatively assess the upper limb motion in the grasping test.

To begin with, an SRF network is designed to recognize rehabilitation target objects grasped in assessment tests. Moreover, the upper limb motion trajectory is calculated through the motion of objects’ central positions. After that, a GAE network is designed to analyze the motion trajectory which reflects the motion of upper limb. Finally, based on the upper limb rehabilitation exoskeleton platform, the upper limb motion assessment tests are carried out to show the accuracy of both object recognition of SRF network and motion assessment of GAE network. The results including object recognition, trajectory calculation and deviation assessment are given with details.

The performance of the proposed networks is validated by experiments that are developed on the upper limb rehabilitation robot. It is implemented by recognizing rehabilitation target objects, calculating the motion trajectory and grading the upper limb motion performance. It illustrates that the networks, including both object recognition and trajectory evaluation, can grade the upper limb motion functionn accurately, where the accuracy is above 95.0% in different grasping tests.

A novel assessment method of upper limb motion is proposed and verified. According to the experimental results, the accuracy can be remarkably enhanced, and the stability of the results can be improved, which provide more quantitative indicators for further application of upper limb motion assessment.

This paper aims to describe a proposal for an innovative method of normal shock wave–turbulent boundary layer interaction (SBLI) and shock-induced separation control.

The concept is based on the introduction of a tangentially moving wall upstream of the shock wave and in the interaction region. The SBLI control mechanism may be implemented as a closed belt floating on an air cushion, sliding over two cylinders and forming the outer skin of the suction side of the airfoil. The presented exploratory numerical study is conducted with SPARC solver (steady 2D RANS). The effect of the moving wall is presented for the NACA 0012 airfoil operating in transonic conditions.

To assess the accuracy of obtained solutions, validation of the computational model is demonstrated against the experimental data of Harris, Ladson & Hill and Mineck & Hartwich (NASA Langley). The comparison is conducted not only for the reference (impermeable) but also for the perforated (permeable) surface NACA 0012 airfoils. Subsequent numerical analysis of SBLI control by moving wall confirms that for the selected velocity ratios, the method is able to improve the shock-upstream boundary layer and counteract flow separation, significantly increasing the airfoil aerodynamic performance.

The moving wall concept as a means of normal shock wave–turbulent boundary layer interaction and shock-induced separation control has been investigated in detail for the first time. The study quantified the necessary operational requirements of such a system and practicable aerodynamic efficiency gains and simultaneously revealed the considerable potential of this promising idea, stimulating a new direction for future investigations regarding SBLI control.

This paper aims to propose a hybrid force/position controller based on the adaptive variable impedance.

First, the working space is divided into a force control subspace and a position subspace, the force control subspace adopts the position impedance control strategy. At the same time, the contact force model between the robot and the surface is analyzed in this space. Second, based on the traditional position impedance, the model reference adaptive control is introduced to provide an accurate reference position for the impedance controller. Then, the BP neural network is used to adjust the impedance parameters online.

The experimental results show that compared with the traditional PI control method, the proposed method has a higher flexibility, the dynamic response accommodation time is reduced by 7.688 s and the steady-state error is reduced by 30.531%. The overshoot of the contact force between the end of robot and the workpiece is reduced by 34.325% comparing with the fixed impedance control method.

The proposed control method compares with a hybrid force/position based on PI control method and a position fixed impedance control method by simulation and experiment.

The adaptive variable impedance control method improves accuracy of force tracking and solves the problem of the large surfaces with robot grinding often over-polished at the protrusion and under-polished at the concave.

Thermal energy storage systems use thermal energy to elevate the temperature of a storage substance, enabling the release of energy during a discharge cycle. The storage or retrieval of energy occurs through the heating or cooling of either a liquid or a solid, without undergoing a phase change, within a sensible heat storage system. In a sensible packed bed thermal energy storage system, the structure comprises porous media that form the packed solid material, while fluid occupies the voids. Thus, a cavity, partially filled with a fluid layer and partially with a saturated porous layer, has become important in the investigation of natural convection heat transfer, carrying significant relevance within thermal energy storage systems. Motivated by these insights, the current investigation delves into the convection heat transfer driven by buoyancy and entropy generation within a partially porous cavity that is differentially heated, vertically layered and filled with a hybrid nanofluid.

The investigation encompasses two distinct scenarios. In the first instance, the porous layer is positioned next to the heated wall, while the opposite region consists of a fluid layer. In the second case, the layers switch places, with the fluid layer adjacent to the heated wall. The system of equations for fluid and porous media, along with appropriate initial and boundary conditions, is addressed using the finite difference method. The Tiwari–Das model is used in this investigation, and the viscosity and thermal conductivity are determined using correlations specific to spherical nanoparticles.

Comprehensive numerical simulations have been performed, considering controlling factors such as the Darcy number, nanoparticle volume fraction, Rayleigh number, bottom slit position and Hartmann number. The visual representation of the numerical findings includes streamlines, isotherms and entropy lines, as well as plots illustrating average entropy generation and the average Nusselt number. These representations aim to provide insight into the influence of these parameters across a spectrum of scenarios.

The computational outcomes indicate that with an increase in the Darcy number, the addition of 2.5% magnetite nanoparticles to the GO nanofluid results in an enhanced heat transfer rate, showing increases of 0.567% in Case 1 and 3.894% in Case 2. Compared with Case 2, Case 1 exhibits a 59.90% enhancement in heat transfer within the enclosure. Positioning the porous layer next to the partially cooled wall significantly boosts the average total entropy production, showing a substantial increase of 11.36% at an elevated Rayleigh number value. Positioning the hot slit near the bottom wall leads to a reduction in total entropy generation by 33.20% compared to its placement at the center and by 33.32% in comparison to its proximity to the top wall.