Simulated Data Research Articles

Balancing Economic Growth and Social Equity: A Legal and Conceptual Analysis of Indonesian Agrarian Reform in the Agribusiness Sector

Indonesia faces the persistent challenge of balancing economic growth with social equity in its agribusiness sector. This study analyzes how legal and conceptual aspects of agrarian reform policies contribute to this balancing act, particularly in the context of rising agribusiness investment and the persistence of smallholder farming. A normative juridical method was employed, incorporating statute, historical, and conceptual approaches. Legal documents, historical records, and academic literature were analyzed. Simulation data, based on government statistics and existing research, was used to model the potential impact of different policy scenarios on economic growth and social equity. Indonesia's legal framework, while promoting land redistribution and tenure security, exhibits inconsistencies that can favor large-scale agribusiness interests. This can lead to unequal access to land and resources, potentially hindering inclusive growth. Simulation data suggests that strengthening land tenure security, improving access to finance, and promoting sustainable agricultural practices can contribute to both economic growth and social equity. In conclusion, optimizing agrarian reform requires a more nuanced approach that prioritizes social equity alongside economic growth. This includes strengthening land rights for smallholder farmers, improving access to resources, and promoting sustainable agricultural practices.

Improved Polar Current Shell Algorithm for Ocean Current Retrieval from X-Band Radar Data

This paper presents an improved algorithm for retrieving ocean surface currents from X-band marine radar images. The original polar current shell (PCS) method begins with a 3D fast Fourier transform (FFT) of the radar image sequence, followed by the extraction of the dispersion shell from the 3D image spectrum, which is then transformed into a PCS using polar coordinates. Building on this foundation, the improved approach is to analyze all data points corresponding to different wavenumber magnitudes in the PCS domain rather than analyzing each specific wavenumber magnitude separately. In addition, kernel density estimation (KDE) to identify high-density directions, interquartile range filtering to remove outliers, and symmetry-based filtering to further reduce noise by comparing data from opposite directions are also utilized for further improvement. Finally, a single curve fitting is applied to the filtered data rather than conducting multiple curve fittings as in the original method. The algorithm is validated using simulated data and real radar data from both the Decca radar, established in 2008, and the Koden radar, established in 2017. For the 2008 Decca radar data, the improved PCS method reduced the root-mean-square deviation (RMSD) for speed estimation by 0.06 m/s and for direction estimation by 3.8° while improving the correlation coefficients (CCs) for current speed by 0.06 and direction by 0.07 compared to the original PCS method. For the 2017 Koden radar data, the improved PCS method reduced the RMSD for speed by 0.02 m/s and for direction by 4.6°, with CCs being improved for current speed by 0.03 and direction by 0.05 compared to the original PCS method.

Isotonic subgroup selection

Abstract Given a sample of covariate–response pairs, we consider the subgroup selection problem of identifying a subset of the covariate domain where the regression function exceeds a predetermined threshold. We introduce a computationally feasible approach for subgroup selection in the context of multivariate isotonic regression based on martingale tests and multiple testing procedures for logically structured hypotheses. Our proposed procedure satisfies a non-asymptotic, uniform Type I error rate guarantee with power that attains the minimax optimal rate up to poly-logarithmic factors. Extensions cover classification, isotonic quantile regression, and heterogeneous treatment effect settings. Numerical studies on both simulated and real data confirm the practical effectiveness of our proposal, which is implemented in the R package ISS.

Dose, exposure, and treatment regimen of intravenous immunoglobulin G in multifocal motor neuropathy

IntroductionIntravenous immunoglobulin (IVIG) is the only approved treatment for multifocal motor neuropathy (MMN), a rare, chronic, immune-mediated demyelinating neuropathy. There is a significant gap in understanding of the role of serum immunoglobulin G (IgG) levels in the efficacy of IVIG in affected patients. We aimed to characterize the interplay between dose and exposure of IVIG and the effects of patient factors on individual variabilities.MethodsSerum IgG trough concentration data from a phase 3, randomized, double-blind, placebo-controlled, crossover trial of IVIG 10% in 44 patients with MMN (NCT00666263) were analyzed using fit-for-purpose population PK modeling. Patient factors were tested as covariates, and IgG PK profiles following various dosing regimens were simulated.ResultsSerum IgG levels, with significant inter-patient variability, correlated with dose and treatment interruptions at the individual patient level. Simulated data for various dosing regimens (0.4–2 g/kg once every 1–4 weeks [Q1–4W]) revealed that more frequent dosing provided more stable IgG levels than less frequent dosing, and dose splitting over multiple days had no significant effects on PK.DiscussionIn patients with MMN, stable dosing and consistent serum IgG levels are crucial to avoid negative responses owing to treatment interruptions. Dosing intervals more frequent than Q4W may alleviate periodic symptom deterioration. Dose splitting potentially offers flexibility for patients requiring large volumes of IVIG without negatively affecting serum IgG PK, while maintaining treatment efficacy. Variability in serum IgG levels between patients suggests that individualizing IVIG treatment regimens and target IgG levels may play a key role in managing MMN.

Contrastive learning-enabled digital twin framework for fault diagnosis of rolling bearing

Abstract Rolling bearings are essential components in various industrial machines, and their failures can lead to significant downtime and maintenance costs. Traditional data-driven fault diagnosis methods often require extensive fault datasets for training, which may not always be available in critical industrial scenarios, limiting their practicality. Digital twins, virtual representations of physical entities reflecting their operational conditions, offer a promising solution for the fault diagnosis of rolling bearings with limited fault data. In this paper, we propose a novel digital twin-driven framework to address the challenge of limited training data in rolling bearing fault diagnosis. Firstly, a virtual bearing simulation model is used to generate the simulated data. Subsequently, a transformer-based network is introduced to learn the discrepancy features from the raw data. Then, a Maximum Mean Discrepancy loss and a supervised contrastive learning loss for raw and augmentation data are established to achieve global domain alignment and instance-based domain alignment. Finally, an unsupervised contrastive learning loss for the augmentation data of the target domain is established to further improve the diagnostic performance. In five cross-domain fault diagnosis tasks representing real industrial scenarios set in this study, the proposed method achieved an average diagnostic accuracy of 84.40%, exceeding two existing advanced domain adaptation methods by more than 10%. The experimental results demonstrate that the proposed method achieves high diagnostic performance in real industrial scenarios where labeled data is lacking. This shows its significant benefits for monitoring the condition of critical bearings.

The recurrent temporal restricted Boltzmann machine captures neural assembly dynamics in whole-brain activity

Animal behaviour alternates between stochastic exploration and goal-directed actions, which are generated by the underlying neural dynamics. Previously, we demonstrated that the compositional Restricted Boltzmann Machine (cRBM) can decompose whole-brain activity of larval zebrafish data at the neural level into a small number (∼100-200) of assemblies that can account for the stochasticity of the neural activity (van der Plas et al., eLife, 2023). Here, we advance this representation by extending to a combined stochastic-dynamical representation to account for both aspects using the recurrent temporal RBM (RTRBM) and transfer-learning based on the cRBM estimate. We demonstrate that the functional advantage of the RTRBM is captured in the temporal weights on the hidden units, representing neural assemblies, for both simulated and experimental data. Our results show that the temporal expansion outperforms the stochastic-only cRBM in terms of generalization error and achieves a more accurate representation of the moments in time. Lastly, we demonstrate that we can identify the original time-scale of assembly dynamics by estimating multiple RTRBMs at different temporal resolutions. Together, we propose that RTRBMs are a valuable tool for capturing the combined stochastic and time-predictive dynamics of large-scale data sets.

CaClust: linking genotype to transcriptional heterogeneity of follicular lymphoma using BCR and exomic variants.

Tumours exhibit high genotypic and transcriptional heterogeneity. Both affect cancer progression and treatment, but have been predominantly studied separately in follicular lymphoma. To comprehensively investigate the evolution and genotype-to-phenotype maps in follicular lymphoma, we introduce CaClust, a probabilistic graphical model integrating deep whole exome, single-cell RNA and B-cell receptor sequencing data to infer clone genotypes, cell-to-clone mapping, and single-cell genotyping. CaClust outperforms a state-of-the-art model on simulated and patient data. In-depth analyses of single cells from four samples showcase effects of driver mutations, follicular lymphoma evolution, possible therapeutic targets, and single-cell genotyping that agrees with an independent targeted resequencing experiment.

The recurrent temporal restricted Boltzmann machine captures neural assembly dynamics in whole-brain activity.

Animal behaviour alternates between stochastic exploration and goal-directed actions, which are generated by the underlying neural dynamics. Previously, we demonstrated that the compositional Restricted Boltzmann Machine (cRBM) can decompose whole-brain activity of larval zebrafish data at the neural level into a small number (∼100-200) of assemblies that can account for the stochasticity of the neural activity (van der Plas et al., eLife, 2023). Here, we advance this representation by extending to a combined stochastic-dynamical representation to account for both aspects using the recurrent temporal RBM (RTRBM) and transfer-learning based on the cRBM estimate. We demonstrate that the functional advantage of the RTRBM is captured in the temporal weights on the hidden units, representing neural assemblies, for both simulated and experimental data. Our results show that the temporal expansion outperforms the stochastic-only cRBM in terms of generalization error and achieves a more accurate representation of the moments in time. Lastly, we demonstrate that we can identify the original time-scale of assembly dynamics by estimating multiple RTRBMs at different temporal resolutions. Together, we propose that RTRBMs are a valuable tool for capturing the combined stochastic and time-predictive dynamics of large-scale data sets.

Toward Accelerating Discovery via Physics-Driven and Interactive Multifidelity Bayesian Optimization

Abstract Both computational and experimental material discovery bring forth the challenge of exploring multidimensional and often nondifferentiable parameter spaces, such as phase diagrams of Hamiltonians with multiple interactions, composition spaces of combinatorial libraries, processing spaces, and molecular embedding spaces. Often these systems are expensive or time consuming to evaluate a single instance, and hence classical approaches based on exhaustive grid or random search are too data intensive. This resulted in strong interest toward active learning methods such as Bayesian optimization (BO) where the adaptive exploration occurs based on human learning (discovery) objective. However, classical BO is based on a predefined optimization target, and policies balancing exploration and exploitation are purely data driven. In practical settings, the domain expert can pose prior knowledge of the system in the form of partially known physics laws and exploration policies often vary during the experiment. Here, we propose an interactive workflow building on multifidelity BO (MFBO), starting with classical (data-driven) MFBO, then expand to a proposed structured (physics-driven) structured MFBO (sMFBO), and finally extend it to allow human-in-the-loop interactive interactive MFBO (iMFBO) workflows for adaptive and domain expert aligned exploration. These approaches are demonstrated over highly nonsmooth multifidelity simulation data generated from an Ising model, considering spin–spin interaction as parameter space, lattice sizes as fidelity spaces, and the objective as maximizing heat capacity. Detailed analysis and comparison show the impact of physics knowledge injection and real-time human decisions for improved exploration with increased alignment to ground truth. The associated notebooks allow to reproduce the reported analyses and apply them to other systems.2

Solid acid-catalyzed green synthesis of bis-Schiff bases: Spectroscopic, DFT, molecular docking, and ADMET studies

This research examines the green synthesis of bis-Schiff bases with benzyl substitutions at 3,4-dimethoxy and 3,4,5-trimethoxy positions through solvent-free microwave-assisted condensation between para-phenylenediamine and 3,4-dimethoxybenzaldehyde (I) and 3,4,5-trimethoxybenzaldehyde (II), using sulfated-TiO2 as a solid acid catalyst. The structures were confirmed through various physicochemical and spectroscopic techniques, including IR, ¹H NMR, ¹³C NMR, and SC-XRD. The compound II crystallized in the monoclinic system (space group of P-21/c). The primary objective was to explore the geometry using both experimental methods and density functional theory (DFT) with the B3LYP/6–311 G (d,p) level. A comparison of experimental (FT-IR, NMR, SC-XRD) and simulated data showed strong agreement. Additionally, Hirshfeld analysis, ELF, LOL, and RDG topological studies were performed. Docking simulations on the 2V54 monkeypox target protein revealed binding affinities, with compounds I and II. The target protein had a binding energy of 4.83 and −4.98 kcal/mol with I and II. ADMET predictions further demonstrated the pharmacological profile of the synthesized compounds.

Dynamic signatures: a mathematical approach to analysis

Abstract This study evaluates mathematical tools (principal component analysis, dynamic time warping, and the Kolmogorov–Smirnov hypothesis test) to analyse global and local data from dynamic signatures to reduce subjectivity and increase the reproducibility of handwriting examination using a two-step approach. A data set composed of 1800 genuine signature samples, 870 simulated signatures, and 60 disguises (30 formally similar or ‘autosimulated’ and 30 random but different from usual) provided by 30 volunteers was collected. The first step involved global data analysis using principal component analysis and a hypothesis test performed for 62 global characteristics, and associations of these characteristics were analysed through calculations of multivariate distance followed by a hypothesis test. The second step involved the analysis of local characteristics including vertical and horizontal positions, speed, pressure gradient, acceleration, and jerk point-to-point, by using dynamic time warping followed by a hypothesis test. Optimization of sensitivity and specificity metrics of the hypothesis test was explored by varying its stringency and observing accuracy rates for the simulated and genuine groups. A P-value threshold of 1 × 10−10 was found to be optimal, making the test more restrictive and yielding accuracy rates of 96.7% for genuine global data and 88.9% for simulated data. The same cutoff value for local characteristics provided an average accuracy rate of 95.4% for genuine samples and 94.7% for simulated samples, demonstrating high accuracy for both simulated and genuine samples. However, the method did not offer reasonable accuracy rates for disguises, consistent with observations in traditional handwriting examination. Our approach provided satisfactory results for forensic examination use. The visualization of graphs and signatures and analysis of all identifying elements of handwriting by the examining expert are still essential. In future studies, we plan to perform blind tests to validate our approach and propose a rigorous methodology.

Data-Driven Remaining Useful Life Estimation

This paper explores the development and application of data-driven prognostic models for estimating the Remaining Useful Life (RUL) of Nuclear Power Plant (NPP) condensers experiencing tube fouling. Due to the unavailability of run-to-failure industry sensor data, we utilized simulated data generated by the Asherah Nuclear Power Plant Simulator (ANS), initially designed by the International Atomic Energy Agency (IAEA) and programmed in Simulink for cyber security simulations. ANS's adaptability allows it to simulate Pressurized Water Reactor (PWR) behaviors given a time series of operating conditions and to introduce degradation modules to mimic fouling effects. Our study compares two primary approaches applied to data generated by ANS: inference-based and direct prediction methods. The selected inference-based approach estimates the health state of the condenser using a pipeline formed by an Auto Associative Kernel Regressor and a Hidden Markov Model (HMM), which subsequently combines the state estimates with its parameters to predict the RUL. The direct prediction method employs a Gradient Boosting Regressor Decision Tree (GBRDT) to map input variables directly to RUL. Our findings demonstrate the efficacy and limitations of each method through the case study, providing valuable insights for the adoption of data-driven RUL estimation techniques in industrial and energy applications.

The structure and dynamics of the laminar separation bubble

A novel selective mode decomposition, proper orthogonal decomposition and dynamic mode decomposition methods are used to analyse large-eddy simulation data of the flow field about a NACA0012 airfoil at low Reynolds numbers of $5\times 10^4$ and $9\times 10^4$ , and at near-stall conditions. The objective of the analysis is to investigate the structure of the laminar separation bubble (LSB) and its associated low-frequency flow oscillation (LFO). It is shown that the flow field can be decomposed into three dominant flow modes: two low-frequency modes (LFO-Mode-1 and LFO-Mode-2) that govern an interplay of a triad of vortices and sustain the LFO phenomenon, and a high-frequency oscillating (HFO) mode featuring travelling Kelvin–Helmholtz waves along the wake of the airfoil. The structure and dynamics of the LSB depend on the energy content of these three dominant flow modes. At angles of attack lower than the stall angle of attack and above the angle of a full stall, the flow is dominated by the HFO mode. At angles of attack above the stall angle of attack the LFO-Mode-2 overtakes the HFO mode, triggers instability in the LSB and initiates the LFO phenomenon. Previous studies peg the structure, stability and bursting conditions of the separation bubble to local flow parameters. However, the amplitude of these local flow parameters is dependent on the energy content of the three dominant flow modes. Thus, the present work proposes a more robust bursting criterion that is based on global eigenmodes.

Temperature management of liquid-cooled fuel cells based on active disturbance rejection control

In the proton exchange membrane fuel cell (PEMFC) systems, the operating temperature affects the gas transport, reaction rate and water balance inside the PEMFC stack, thus affecting the output performance and lifetime of the stack. Therefore, it is necessary to study the thermal management methods and control strategies to maintain the stack temperature at the expected value. A thermal management system model of the PEMFC based on the electrochemical reaction and thermodynamics is established. This study proposes a novel active disturbance rejection controller (ADRC) to solve the problem of excessive temperature fluctuation of liquid-cooled stack with the dynamic load current changes. The simulation data show that compared with the PID, Fuzzy-PID and PSO-PID controllers for the cooling fan, the overshoot amount of the coolant inlet temperature with the ADRC is significantly reduced from 1.02 %, 0.77 % and 0.26 % to 0.07 %, and the integral of time absolute error (ITAE) is decreased from 5.299 × 104, 5.206 × 104 and 1.255 × 104 to 3.930 × 103, respectively. The degree of temperature fluctuation is reduced and the parasitic power does not significantly increase. Therefore, the ADRC proposed in this study improves the temperature control effect of the PEMFC stack, which is conducive to enhancing the stack performance.

An improved transfer learning method for rotor unbalance position identification from simulated data to experimental data

In the study of dynamic balancing for flexible rotors operating at high speeds, determining the unbalanced position of the rotor has consistently posed significant challenges. Accurate identification of the unbalanced position enables low-speed dynamic balancing to serve as a viable alternative to high-speed methods, ultimately reducing costs and enhancing operational efficiency. Although deep learning techniques utilizing limited labeled data have shown promising results in identifying unbalanced positions, the challenge of gathering a sufficient amount of labeled data for effectively training diagnostic models remains substantial. To address this issue, we propose a Pre-Adaptive Transfer Learning (PATL) approach that employs sample reconstruction through frequency domain correlation analysis. This technique facilitates cross-domain deep transfer recognition of rotor unbalanced positions by transferring insights from simulated dynamic model data to experimental datasets. Compared to existing methodologies, the proposed approach demonstrates significant improvements in both the accuracy and generalization of unbalanced position identification. The novelty of this research lies in the introduction of pre-adaptive transfer learning, which effectively minimizes the disparity between experimental and simulation data, thereby enhancing the model's recognition capabilities. Experimental results indicate that the proposed method achieves effective balancing outcomes across various operating conditions.

Mapping eccentricity evolutions between numerical relativity and effective-one-body gravitational waveforms

Orbital eccentricity in compact binaries is considered to be a key tracer of their astrophysical origin, and can be inferred from gravitational-wave observations due to its imprint on the emitted signal. For a robust measurement, accurate waveform models are needed. However, ambiguities in the definition of eccentricity can obfuscate the physical meaning and result in seemingly discrepant measurements. In this work we present a suite of 28 new numerical relativity simulations of eccentric, aligned-spin binary black holes with mass ratios between 1 and 6 and initial post-Newtonian eccentricities between 0.05 and 0.3. We then develop a robust pipeline for measuring the eccentricity evolution as a function of frequency from gravitational-wave observables that is applicable even to signals that span at least ≳7 orbits. We assess the reliability of our procedure and quantify its robustness under different assumptions on the data. Using the eccentricity measured at the first apastron, we initialize effective-one-body waveforms and quantify how the precision in the eccentricity measurement, and therefore the choice of the initial conditions, impacts the agreement with the numerical data. We find that even small deviations in the initial eccentricity can lead to non-negligible differences in the phase and amplitude of the waveforms. However, we demonstrate that we can reliably map the eccentricities between the simulation data and analytic models, which is crucial for robustly building eccentric hybrid waveforms, and to improve the accuracy of eccentric waveform models in the strong-field regime. Published by the American Physical Society 2024

Study on the accurate detection method of full‐polarimetric ground‐penetrating radar faults in mines based on modified Yamaguchi decomposition

AbstractObtaining information on tectonic tendencies is a prerequisite for intelligent and accurate mining in mines. In the special mine environment, the co‐polarized ground‐penetrating radar can only identify the spatial location of faults, and it is difficult to analyse the inclination information of fault structures. This paper proposes a mine full‐polarimetric ground‐penetrating radar fault tendency detection method based on this. First, based on the stacking characteristics of the coal depositional, this paper analyses the propagation law of the pulse electromagnetic wave in the coal seam and puts forward the assumption of the overlapping echo reflection of the fault structure. The reasonableness of the fault reflection assumption is verified through a numerical simulation study. Second, based on the cutting relationship of the fault to the coal seam, we divided the reflection structure of the fault structure into plane scattering and dihedral angle scattering. We realized the mingled echoes’ decomposition using the improved Yamaguchi decomposition technique. To analyse the applicability of the modified Yamaguchi and Freeman decomposition methods in the identification of fault inclination, we use the upright fault simulation data for the discussion, and we find that the improved Yamaguchi decomposition method is more advantageous in the identification of fault inclination in the mine. The decomposition results based on the simulation data of fault models with different dip angles found that when the dip angle of the fault is less than 90°, the scattering of the fault structure is dominated by planar scattering and dihedral angle scattering; when the dip angle of the fault is greater than or equal to 90°, the scattering of the fault structure is dominated by planar scattering, and the scattering power of the dihedral angle model is zero. By analysing the effect of fault strike on the decomposition results, it is found that the fault strike angle has little effect on the identification of fault tendency. Finally, the application potential of this paper's method is tested by constructing complex numerical models and probing experiments. Therefore, the method proposed in this paper can solve the fault tendency identification under a thick coal seam.

Numerical and Experimental Analysis of Roller Hemming on Door Panel’s Curved and Straight-Edge of Flat Plane

Owing to its enhanced production efficiency, roller hemming has become the mainstream process for forming and joining metal sheets in the automotive industry. This study investigates the roller hemming process of a specific car door panel through a combination of experimental analysis and finite element analysis (FEA) on both straight-edge and curved-edge flat surfaces. Consequently, the mechanical properties of the door panel, including tensile strength, yield strength, modulus of elasticity, and Poisson’s ratio, were estimated through tensile testing and then underwent finite element modeling. The simulation results demonstrated the varying distribution of stress during the rolling hemming process, with the highest stress concentration observed in the bending area. Additionally, creepage and growing results were acquired from both simulation and experimental data to validate the precision of the numerical model. A comparison was made between the experimental and simulation results of the external forces exerted by the roller on the panel. In both straight- and curved-edge sections, the external force during final hemming exceeded that during pre-hemming, as revealed by experimental measurements of both normal and tangential external forces, surpassing their corresponding simulated values.

Terraces in species tree inference from gene trees

A terrace in a phylogenetic tree space is a region where all trees contain the same set of subtrees, due to certain patterns of missing data among the taxa sampled, resulting in an identical optimality score for a given data set. This was first investigated in the context of phylogenetic tree estimation from sequence alignments using maximum likelihood (ML) and maximum parsimony (MP). It was later extended to the species tree inference problem from a collection of gene trees, where a set of equally optimal species trees was referred to as a “pseudo” species tree terrace which does not consider the topological proximity of the trees in terms of the induced subtrees resulting from certain patterns of missing data. In this study, we mathematically characterize species tree terraces and investigate the mathematical properties and conditions that lead multiple species trees to induce/display an identical set of locus-specific subtrees owing to missing data. We report that species tree terraces are agnostic to gene tree heterogeneity. Therefore, we introduce and characterize a special type of gene tree topology-aware terrace which we call “peak terrace”. Moreover, we empirically investigated various challenges and opportunities related to species tree terraces through extensive empirical studies using simulated and real biological data. We demonstrate the prevalence of species tree terraces and the resulting ambiguity created for tree search algorithms. Remarkably, our findings indicate that the identification of terraces could potentially lead to advances that enhance the accuracy of summary methods and provide reasonably accurate branch support.

Effects of atmospheric pressure change during flight on insulin pump delivery and glycaemic control of pilots with insulin-treated diabetes: an in vitro simulation and a retrospective observational real-world study.

Glycaemic control and clinical outcomes in diabetes are improved by continuous subcutaneous insulin infusion (CSII). Atmospheric pressure changes during flights may affect insulin delivery from pumps and cause unintended metabolic consequences, including hypoglycaemia, in people with type 1 diabetes. The present report evaluates both hypobaric flight simulation and real-world data in pilots using insulin pumps while flying. In the flight simulation part of this study, an in vitro study of insulin pumps was conducted in a hypobaric chamber, de-pressurised to 550 mmHg to mimic the atmospheric pressure changes in airliner cabins during commercial flights. Insulin delivery rates and bubble formation were recorded for standard flight protocol. Insulin infusion sets, without pumps, were tested in a simulated rapid decompression scenario. The real-world observational study was a 7.5-year retrospective cohort study in which pre- and in-flight self-monitored blood glucose (SMBG) values were monitored in pilots with insulin-treated diabetes. Commercial and private pilots granted a medical certificate to fly within the European Union Aviation Safety Agency approved protocol and receiving insulin either by pump or multiple daily injections (MDI) were included. In the flight simulation study, full cartridges over-delivered 0.60 U of insulin during a 20 min ascent and under-delivered by 0.51 U during descent compared with ground-level performance. During emergency rapid decompression, 5.6 U of excess insulin was delivered. In the real-world study, seven pilots using CSII recorded 4656 SMBG values during 2345 h of flying across 1081 flights. Only 33 (0.7%) values were outside an acceptable safe range (5.0-15.0 mmol/l [90-270 mg/dl]). No clinically significant fall in the median SMBG concentration was observed after aircraft ascent and no in-flight SMBG values were within the hypoglycaemic range (<4.0 mmol/l [<72 mg/dl]). Compared with pilots receiving MDI therapy, pilots using CSII recorded more SMBG values within the acceptable range (99.3% vs 97.5%), fewer values in the low red range (0.02% vs 0.1%), fewer in-flight out-of-range values (0.2% vs 1.3%) and maintained stricter glycaemic control during flight. Ambient pressure reduction during simulated flights results in bubble formation and expansion within insulin cartridges. This causes unintended delivery of small insulin doses independent of pre-determined delivery rates and represents the maximum amount of insulin that could be delivered and retracted. However, in vivo, pilots using CSII in-flight did not experience a fall in blood glucose or episodes of hypoglycaemia during these atmospheric pressure changes and the use of insulin pumps can be endorsed in view of their clinical benefits.