Search

High-precision surface defect detection in manufacturing is essential for ensuring quality control. Laser triangulation profilometric sensors are key to this process, providing detailed and accurate surface measurements over a line. To achieve a complete and precise surface scan, accurate relative motion between the sensor and the workpiece is required. It is crucial to control the sensor pose to maintain optimal distance and relative orientation to the surface. It is also important to ensure uniform profile distribution throughout the scanning process. This paper presents a novel Reinforcement Learning (RL) based approach to optimize robot inspection trajectories for profilometric sensors. Building upon the Boustrophedon scanning method, our technique dynamically adjusts the sensor position and tilt to maintain optimal orientation and distance from the surface, while also ensuring a consistent profile distance for uniform and high-quality scanning. Utilizing a simulated environment based on the CAD model of the part, we replicate real-world scanning conditions, including sensor noise and surface irregularities. This simulation-based approach enables offline trajectory planning based on CAD models. Key contributions include the modeling of the state space, action space, and reward function, specifically designed for inspection applications using profilometric sensors. We use Proximal Policy Optimization (PPO) algorithm to efficiently train the RL agent, demonstrating its capability to optimize inspection trajectories with profilometric sensors. To validate our approach, we conducted several experiments where a model trained on a specific training piece was tested on various parts in simulation. Also, we conducted a real-world experiment by executing the optimized trajectory, generated offline from a CAD model, to inspect a part using a UR3e robotic arm model.

Large language models (LLMs) hold potential for innovative HCI research, including the creation of synthetic personae. However, their black-box nature and propensity for hallucinations pose challenges. To address these limitations, this position paper advocates for using LLMs as data augmentation systems rather than zero-shot generators. We further propose the development of robust cognitive and memory frameworks to guide LLM responses. Initial explorations suggest that data enrichment, episodic memory, and self-reflection techniques can improve the reliability of synthetic personae and open up new avenues for HCI research., Comment: This paper was accepted for publication: Proceedings of ACM Conf on Human Factors in Computing Systems (CHI 24), Rafael Arias Gonzalez, Steve DiPaola. Exploring Augmentation and Cognitive Strategies for Synthetic Personae. ACM SigCHI, in Challenges and Opportunities of LLM-Based Synthetic Personae and Data in HCI Workshop, 2024

We consider a class of first-order partial differential operators, acting on the space of ultradifferentiable periodic functions, and we describe their range by using the following conditions on the coefficients of the operators: the connectedness of certain sublevel sets, the dimension of the subspace generated by the imaginary part of the coefficients, and Diophantine conditions. In addition, we show that these properties are also linked to the regularity of the solutions. The results extend previous ones in Gevrey classes.

Although a catalogue of synthetic RGB magnitudes, providing photometric data for a sample of 1346 bright stars, has been recently published, its usefulness is still limited due to the small number of reference stars available, considering that they are distributed throughout the whole celestial sphere, and the fact that they are restricted to Johnson V < 6.6 mag. This work presents synthetic RGB magnitudes for ~15 million stars brighter than Gaia G = 18 mag, making use of a calibration between the RGB magnitudes of the reference bright star sample and the corresponding high quality photometric G, G_BP and G_RP magnitudes provided by the Gaia EDR3. The calibration has been restricted to stars exhibiting -0.5 < G_BP - G_RP < 2.0 mag, and aims to predict RGB magnitudes within an error interval of $\pm 0.1$ mag. Since the reference bright star sample is dominated by nearby stars with slightly undersolar metallicity, systematic variations in the predictions are expected, as modelled with the help of stellar atmosphere models. These deviations are constrained to the $\pm 0.1$ mag interval when applying the calibration only to stars scarcely affected by interstellar extinction and with metallicity compatible with the median value for the bright star sample. The large number of Gaia sources available in each region of the sky should guarantee high-quality RGB photometric calibrations., Comment: 12 pages, 9 figures, 2 tables. Accepted for publication in MNRAS

Although the use of RGB photometry has exploded in the last decades due to the advent of high-quality and inexpensive digital cameras equipped with Bayer-like color filter systems, there is surprisingly no catalogue of bright stars that can be used for calibration purposes. Since due to their excessive brightness, accurate enough spectrophotometric measurements of bright stars typically cannot be performed with modern large telescopes, we have employed historical 13-color medium-narrow-band photometric data, gathered with quite reliable photomultipliers, to fit the spectrum of 1346 bright stars using stellar atmosphere models. This not only constitutes a useful compilation of bright spectrophotometric standards well spread in the celestial sphere, the UCM library of spectrophotometric spectra, but allows the generation of a catalogue of reference RGB magnitudes, with typical random uncertainties $\sim 0.01$ mag. For that purpose, we have defined a new set of spectral sensitivity curves, computed as the median of 28 sets of empirical sensitivity curves from the literature, that can be used to establish a standard RGB photometric system. Conversions between RGB magnitudes computed with any of these sets of empirical RGB curves and those determined with the new standard photometric system are provided. Even though particular RGB measurements from single cameras are not expected to provide extremely accurate photometric data, the repeatability and multiplicity of observations will allow access to a large amount of exploitable data in many astronomical fields, such as the detailed monitoring of light pollution and its impact on the night sky brightness, or the study of meteors, solar system bodies, variable stars, and transient objects. In addition, the RGB magnitudes presented here make the sky an accessible and free laboratory for the calibration of the cameras themselves., Comment: 19 pages, 18 figures, 7 tables. Accepted for publication in MNRAS

Nanoparticles are ubiquitous in nature and are increasingly important for technology. They are subject to bombardment by ionizing radiation in a diverse range of environments. In particular, nanodiamonds represent a variety of nanoparticles of significant fundamental and applied interest. Here we present a combined experimental and computational study of the behaviour of nanodiamonds under irradiation by xenon ions. Unexpectedly, we observed a pronounced size effect on the radiation resistance of the nanodiamonds: particles larger than 8 nm behave similarly to macroscopic diamond (i.e. characterized by high radiation resistance) whereas smaller particles can be completely destroyed by a single impact from an ion in a defined energy range. This latter observation is explained by extreme heating of the nanodiamonds by the penetrating ion. The obtained results are not limited to nanodiamonds, making them of interest for several fields, putting constraints on processes for the controlled modification of nanodiamonds, on the survival of dust in astrophysical environments, and on the behaviour of actinides released from nuclear waste into the environment., Comment: Supplementray information is at: doi:10.1038/s41598-018-23434-y