Institute of Intelligent Systems for Automation

Waste wool hydrolysates (WWHs), a by‐product originating from the alkaline hydrolysis of waste wool, are recovered and employed as auxiliaries in wool dyeing. In view of an eco‐friendly dyeing procedure, a natural dye, Carmine, is selected to dye wool fabrics. Different methodologies for performing the dyeing process are described. In the first procedure, the wool fabrics are pretreated with a water suspension of the WWHs at room temperature, left overnight, and then cured at 180 °C. In the second procedure, the wool fabrics are immersed in the WWH's suspension at 100 °C, dried in an oven, and subsequently dyed through the exhaustion method. In the last procedure, the WWHs are added directly to the dyeing liquor. Dye exhaustion, color coordinates, and K/S are measured to evaluate the dyeing efficiency. The dyed fabrics are also characterized in terms of thermal, chemical, mechanical and morphological properties. The results demonstrate that the WWHs are efficient alternatives to metal‐based mordants in assisting wool dyeing with Carmine dye. The evidence of non‐significant damages to fabrics as a consequence of the chosen treatment conditions further supports the possibility of WWH valorization in textile industries as a by‐product that otherwise would represent a waste to dispose of.

Cup-like nuclear morphological alterations in acute myeloid leukemia (AML) blasts have been widely correlated with Nucleophosmin 1 (NPM1) mutations. NPM1-mutated AML has earned recognition as a distinct entity among myeloid tumors, but the absence of a thoroughly established tool for its morphological analysis remains a notable gap. Holographic tomography (HT) can offer a label-free solution for quantitatively assessing the 3D shape of the nucleus based on the volumetric variations of its refractive indices (RIs). However, traditional HT methods analyze adherent cells in a 2D layer, leading to non-isotropic reconstructions due to missing cone artifacts. Here we show for the first time that holo-tomographic flow cytometry (HTFC) achieves quantitative specificity and precise capture of the nucleus volumetric shape in AML cells in suspension. To retrieve nucleus specificity in label-free RI tomograms of flowing AML cells, we conceive and demonstrate in a real-world clinical case a novel strategy for segmenting 3D concave nuclei. This method implies that the correlation between the “phenotype” and “genotype” of nuclei is demonstrated through HTFC by creating a challenging link not yet explored between the aberrant morphological features of AML nuclei and NPM1 mutations. We conduct an ensemble-level statistical characterization of NPM1-wild type and NPM1-mutated blasts to discern their complex morphological and biophysical variances. Our findings suggest that characterizing cup-like nuclei in NPM1-related AML cells by HTFC may enhance the diagnostic approach for these tumors. Furthermore, we integrate virtual reality to provide an immersive fruition of morphological changes in AML cells within a true 3D environment.

Ground Penetrating Radar (GPR) has become a valuable tool in precision agriculture for subsurface analysis. This work presents the adaptation of GPR to an autonomous robot that is being designed for agricultural tasks, capable of mapping the terrain using both Lidar and Computer Vision (CV) technologies. The integration of these systems will allow the robot to navigate autonomously while collecting data for mapping the surface and subsurface of the soil. The study focuses on the use of GPR data detect changes in the water content, enabling the robot to create detailed moisture maps of the shallow subsurface, providing insights into soil hydration levels. An additional objective is to enable the data collected by the robot to generate root maps, essential for understanding root growth and distribution. To achieve this, the aim is to determine the most effective method for detecting roots using GPR. The initial results of the project highlight the potential of autonomous robots equipped with GPR for optimizing agricultural practices, enhancing efficiency in field analysis and improving decision-making processes.

In the Industry 5.0 era, the focus shifts from basic automation to fostering collaboration between humans and robots. Trust is crucial in this new paradigm, enabling smooth interaction, especially for users with limited robotics knowledge. This study presents a novel framework that uses human hand gestures and voice commands to control robot movements, aiming to enhance trust, reduce cognitive workload, and minimize task execution time-key for efficient manufacturing. In automated systems, swift completion of micromanagement tasks is essential to prevent process disruption. To evaluate this framework, we devised a testbed scenario within an automated carbon fiber transportation and draping process, focusing on a maintenance task as the micromanagement challenge. Participants inspected the gripper, guided the robot along a defined path, and performed maintenance, such as attaching cables. Two conditions were tested: gestures and voice commands versus a smartPAD. The results showed that gestures and voice commands increased trust, lowered cognitive load, and shortened execution times, improving overall manufacturing efficiency.

The explosivity of a volcanic eruption is controlled by several interdependent processes during magma ascent, such as crystallisation, gas exsolution and outgassing. Syn-eruptive crystallisation can increase the potential of magma fragmentation. Whilst the degree of coupling between the gas and melt phases during ascent can influence eruptive style. Quantitative textural analysis of vesicles and crystals in erupted products can provide insight into syn-eruptive conduit processes and the conditions leading to magma fragmentation. Synchrotron-based imaging techniques such as X-ray computed micro-tomography can provide information on vesicle and crystal size, shape and their spatial distribution in 3D. Furthermore, X-ray ptychography, an X-ray microscopy technique with nanoscale resolution, can be used to expand this 3D textural analysis to nanoscale crystals in volcanic rocks. Here, we present a 3D reconstruction and quantification of vesicle and crystal textures in pyroclasts of the Masaya Triple Layer eruption, a highly explosive Plinian eruption of Masaya caldera, Nicaragua. Images and observations of vesicle textures at the micro-scale were acquired using X-ray computed micro-tomography and used to reconstruct the geometrical properties of the connected pore network, including connectivity, tortuosity and the throat-pore size ratio. X-ray ptychography was used to perform a 3D textural analysis of nanoscale crystals within the groundmass of clasts. These data were used to reconstruct conduit processes and evaluate the impact of syn-eruptive crystallisation, vesiculation and outgassing on magma rheology and fragmentation. Our results provide insight into the driving mechanisms of highly explosive, basaltic Plinian activity, and also highlight the potential of using multi-scale 3D imaging techniques to analyse textural features in pyroclasts and investigate controls on eruptive style.

Latest advances in artificial intelligence, particularly in object recognition and segmentation, provide unprecedented opportunities for precision agriculture. This work investigates the use of state-of-the-art AI models, namely Meta’s Segment Anything (SAM) and GroundingDino, for the task of grape cluster detection in vineyards. Three different methods aimed at enhancing the instance segmentation process are proposed: (i) SAM-Refine (SAM-R), which refines a previously proposed depth-based clustering approach, referred to as DepthSeg, using SAM; (ii) SAM-Segmentation (SAM-S), which integrates SAM with a pre-trained semantic segmentation model to improve cluster separation; and (iii) AutoSAM-Dino (ASD), which eliminates the need for manual labeling and transfer learning through the combined use of GroundingDino and SAM. Analysis is conducted on both the object counting and pixel-level segmentation accuracy against a manually labeled ground truth. Metrics such as mean Average Precision (mAP), Intersection over Union (IoU), and precision and recall are calculated to assess the system performance. Compared to the original DepthSeg algorithm, SAM-R slightly advances object counting (mAP: +0.5%) and excels in pixel-level segmentation (IoU: +17.0%). SAM-S, despite a mAP decrease, improves segmentation accuracy (IoU: +13.9%, Precision: +9.2%, Recall: +11.7%). Similarly, ASD, although with a lower mAP, shows significant accuracy enhancement (IoU: +7.8%, Precision: +4.2%, Recall: +4.9%). Additionally, from a labor effort point of view, instance segmentation techniques require much less time for training than manual labeling.

In this chapter, we not only highlight the great potential of EO data and the achievements of recent remote sensing technologies, but also their challenges and limitations, in support of biodiversity and ecosystem monitoring. In Europe, nature conservation rests upon a strong yet ambitious policy framework with legally binding directives. Also, in other parts of the world, environmental legislation follows ambitious goals that often have to compete with other societal premises, such as growth, production, and expansion. Thus, geospatial information products are required at all levels of implementation. With recent advances in EO data availability and forthcoming capable analysis tools, we enter a new dimension of satellite-based information services. Recent technological advancements in this application domain are discussed, using spearheading examples from inside and outside Europe.