arXiv subjects

P. Sánchez

Publications and source records attributed to P. Sánchez.

7 recordsLinked to original sources

arXiv2026

Assesing the Viability of Unsupervised Learning with Autoencoders for Predictive Maintenance in Helicopter Engines

Unplanned engine failures in helicopters can lead to severe operational disruptions, safety hazards, and costly repairs. To mitigate these risks, this study compares two predictive maintenance strategies for helicopter engines: a supervised classification pipeline and an unsupervised anomaly detection approach based on autoencoders (AEs). The supervised method relies on labelled examples of both normal and faulty behaviour, while the unsupervised approach learns a model of normal operation using only healthy engine data, flagging deviations as potential faults. Both methods are evaluated on a real-world dataset comprising labelled snapshots of helicopter engine telemetry. While supervised models demonstrate strong performance when annotated failures are available, the AE achieves effective detection without requiring fault labels, making it particularly well suited for settings where failure data are scarce or incomplete. The comparison highlights the practical trade-offs between accuracy, data availability, and deployment feasibility, and underscores the potential of unsupervised learning as a viable solution for early fault detection in aerospace applications.

cs.LG↗

arXiv2026

LSTM VS. Feed-Forward Autoencoders for Unsupervised Fault Detection in Hydraulic Pumps

Unplanned failures in industrial hydraulic pumps can halt production and incur substantial costs. We explore two unsupervised autoencoder (AE) schemes for early fault detection: a feed-forward model that analyses individual sensor snapshots and a Long Short-Term Memory (LSTM) model that captures short temporal windows. Both networks are trained only on healthy data drawn from a minute-level log of 52 sensor channels; evaluation uses a separate set that contains seven annotated fault intervals. Despite the absence of fault samples during training, the models achieve high reliability.

cs.LG↗

arXiv2026

Early Fault Detection on CMAPSS with Unsupervised LSTM Autoencoders

This paper introduces an unsupervised health-monitoring framework for turbofan engines that does not require run-to-failure labels. First, operating-condition effects in NASA CMAPSS sensor streams are removed via regression-based normalisation; then a Long Short-Term Memory (LSTM) autoencoder is trained only on the healthy portion of each trajectory. Persistent reconstruction error, estimated using an adaptive data-driven threshold, triggers real-time alerts without hand-tuned rules. Benchmark results show high recall and low false-alarm rates across multiple operating regimes, demonstrating that the method can be deployed quickly, scale to diverse fleets, and serve as a complementary early-warning layer to Remaining Useful Life models.

cs.LG↗

arXiv2024

Physical properties of asteroid Dimorphos as derived from the DART impact

On September 26, 2022, NASA's Double Asteroid Redirection Test (DART) mission successfully impacted Dimorphos, the natural satellite of the binary near-Earth asteroid (65803) Didymos. Numerical simulations of the impact provide a means to explore target surface material properties and structures, consistent with the observed momentum deflection efficiency, ejecta cone geometry, and ejected mass. Our simulation, which best matches observations, indicates that Dimorphos is weak, with a cohesive strength of less than a few pascals (Pa), similar to asteroids (162173) Ryugu and (101955) Bennu. We find that a bulk density of Dimorphos, rhoB, lower than 2400 kg/m3, and a low volume fraction of boulders (<40 vol%) on the surface and in the shallow subsurface, are consistent with measured data from the DART experiment. These findings suggest Dimorphos is a rubble pile that might have formed through rotational mass shedding and re-accumulation from Didymos. Our simulations indicate that the DART impact caused global deformation and resurfacing of Dimorphos. ESA's upcoming Hera mission may find a re-shaped asteroid, rather than a well-defined crater.

astro-ph.EP↗

arXiv2019

Small Solar System Bodies as granular media

Asteroids and other Small Solar System Bodies (SSSBs) are of high general and scientific interest in many aspects. The origin, formation, and evolution of our Solar System (and other planetary systems) can be better understood by analysing the constitution and physical properties of small bodies in the Solar System. Currently, two space missions (Hayabusa2, OSIRIS-REx) have recently arrived at their respective targets and will bring a sample of the asteroids back to Earth. Other small body missions have also been selected by, or proposed to, space agencies. The threat posed to our planet by near-Earth objects (NEOs) is also considered at the international level, and this has prompted dedicated research on possible mitigation techniques. The DART mission, for example, will test the kinetic impact technique. Even ideas for industrial exploitation have risen during the last years. Lastly, the origin of water and life on Earth appears to be connected to asteroids. Hence, future space mission projects will undoubtedly target some asteroids or other SSSBs. In all these cases and research topics, specific knowledge of the structure and mechanical behaviour of the surface as well as the bulk of those celestial bodies is crucial. In contrast to large telluric planets and dwarf planets, a large proportion of such small bodies is believed to consist of gravitational aggregates ('rubble piles') with no -- or low -- internal cohesion, with varying macro-porosity and surface properties (from smooth regolith covered terrain, to very rough collection of boulders), and varying topography (craters, depressions, ridges) [...].

astro-ph.EP↗

arXiv2018

Type II Supernovae as Distance Indicators at Near-IR Wavelengths

Motivated by the advantages of observing at near-IR wavelengths, we investigate Type II supernovae (SNe II) as distance indicators at those wavelengths through the Photospheric Magnitude Method (PMM). For the analysis, we use $BVIJH$ photometry and optical spectroscopy of 24 SNe II during the photospheric phase. To correct photometry for extinction and redshift effects, we compute total-to-selective broadband extinction ratios and $K$-corrections up to $z=0.032$. To estimate host galaxy colour excesses, we use the colour-colour curve method with the $V\!-\!I$ versus $B\!-\!V$ as colour combination. We calibrate the PMM using four SNe II in galaxies having Tip of the Red Giant Branch distances. Among our 24 SNe II, nine are at $cz>2000$ km s$^{-1}$, which we use to construct Hubble diagrams (HDs). To further explore the PMM distance precision, we include into HDs the four SNe used for calibration and other two in galaxies with Cepheid and SN Ia distances. With a set of 15 SNe II we obtain a HD rms of 0.13 mag for the $J$-band, which compares to the rms of 0.15-0.26 mag for optical bands. This reflects the benefits of measuring PMM distances with near-IR instead of optical photometry. With the evidence we have, we can set the PMM distance precision with $J$-band below 10 per cent with a confidence level of 99 per cent.

astro-ph.CO↗

arXiv2017

Near Infrared Variability of obscured and unobscured X-ray selected AGN in the COSMOS field

We present our statistical study of near infrared (NIR) variability of X-ray selected Active Galactic Nuclei (AGN) in the COSMOS field, using UltraVISTA data. This is the largest sample of AGN light curves in YJHKs bands, making possible to have a global description of the nature of AGN for a large range of redshifts, and for different levels of obscuration. To characterize the variability properties of the sources we computed the Structure Function. Our results show that there is an anti-correlation between the Structure Function $A$ parameter (variability amplitude) and the wavelength of emission, and a weak anti-correlation between $A$ and the bolometric luminosity. We find that Broad Line (BL) AGN have a considerably larger fraction of variable sources than Narrow Line (NL) AGN, and that they have different distributions of the $A$ parameter. We find evidence that suggests that most of the low luminosity variable NL sources correspond to BL AGN, where the host galaxy could be damping the variability signal. For high luminosity variable NL, we propose that they can be examples of "True type II" AGN or BL AGN with limited spectral coverage which results in missing the Broad Line emission. We also find that the fraction of variable sources classified as unobscured in the X-ray is smaller than the fraction of variable sources unobscured in the optical range. We present evidence that this is related to the differences in the origin of the obscuration in the optical and X-ray regimes.

astro-ph.GA↗