G. M. Sacha

Artificial intelligence in nanotechnology

During the last decade there has been increasing use of artificial intelligence tools in nanotechnology research. In this paper we review some of these efforts in the context of interpreting scanning probe microscopy, the study of biological nanosystems, the classification of material properties at the nanoscale, theoretical approaches and simulations in nanoscience, and generally in the design of nanodevices. Current trends and future perspectives in the development of nanocomputing hardware that can boost artificial-intelligence-based applications are also discussed. Convergence between artificial intelligence and nanotechnology can shape the path for many technological developments in the field of information sciences that will rely on new computer architectures and data representations, hybrid technologies that use biological entities and nanotechnological devices, bioengineering, neuroscience and a large variety of related disciplines.

Influence of the macroscopic shape of the tip on the contrast in scanning polarization force microscopy images

We demonstrate that a quantitative analysis of the contrast obtained in electrostatic force microscopy images that probe the dielectric response of the sample (scanning polarization force microscopy (SPFM)) requires numerical simulations that take into account both the macroscopic shape of the tip and the nanoscopic tip apex. To simulate the SPFM contrast, we have used the generalized image charge method (GICM), which is able to accurately deal with distances between a few nanometers and several microns, thus involving more than three orders of magnitude. Our numerical simulations show that the macroscopic shape of the tip accounts for most of the SPFM contrast. Moreover, we find a quasi-linear relation between the working tip-sample distance and the contrast for tip radii between 50 and 200 nm. Our calculations are compared with experimental measurements of the contrast between a thermally grown silicon oxide sample and a few-layer graphene film transferred onto it.

Generalized Image Charge Method to Calculate Electrostatic Magnitudes at the Nanoscale Powered by Artificial Neural Networks

A technique to calculate electrostatic magnitudes such as force and potential in Electrostatic Force Microscopy setups is presented. This technique combines Artificial Neural Networks and the Generalized Image Charge Method to overcome one of the main problems of traditional numerical simulations: the need of many parameters that are difficult to estimate and depend on the geometry of the experimental setup. Using Artificial Neural Networks, our technique is able to estimate the internal parameters of the algorithm and automatically obtain the electric magnitudes with a very high accuracy. This technique has been implemented in the freely distributed software winGICM. The automatic configuration of the software by an Artificial Neural Network allows the users to handle it without being specifically trained in the theoretical background underlying the algorithms.

Influence of the Substrate and Tip Shape on the Characterization of Thin Films by Electrostatic Force Microscopy

Electrostatic force microscopy has been shown to be a useful tool to determine the dielectric constant of nanoscaled thin films that play a key role in many electrical, optical, and biological phenomena. Previous approaches have made use of simple analytical models to analyze the experimental data for these materials. Here, we show that the electrostatic force shows a completely different behavior when the shape of the tip and sample are taken into account. We present a complete study of the interaction between the whole tip and the layers below the thin film. We demonstrate that physical magnitudes such as the surface charge density distribution and the size of the materials have a strong influence on the EFM signal. The EFM sensitivity to the substrate below the thin film decreases with the substrate thickness and saturates for thicknesses above two times the length of the tip, when it is close to that of an infinite medium.

Computer-assisted assessment with item classification for programming skills

In this work, we propose a computer-assisted method to improve the quality of the assessment process for subjects related to basic programming. We demonstrate, through numerical simulations and experiments with real students, that our models can be very useful when compared with traditional evaluation processes. The characteristics of our method improve the objectiveness, security and take into account the relevance of the subject contents. Results from this work can be directly used in computer-assisted tests for different subjects and disciplines. Even, some self-evaluation tools can be developed to be used by the students with the objective of correcting their deficiencies in the learning process if needed.

eMadrid project: Authoring, reuse and remote labs

This article describes in detail the main achievements in authoring and reuse of educational materials within the eMadrid project.

eMadrid project: Ubiquitous learning, adaptation, adaptability and accessibility

In this paper we present the studies and research conducted within the eMadrid project, funded by the Regional Government of Comunidad Autónoma de Madrid. In particular, we focus on those works dealing with two research lines: ubiquitous and mobile learning; adaptation, adaptability and accessibility.