Gross morphology and adhesion-associated physical properties of Drosophila larval salivary gland glue secretion.

One of the major functions of the larval salivary glands (SGs) of many Drosophila species is to produce a massive secretion during puparium formation. This so-called proteinaceous glue is exocytosed into the centrally located lumen, and subsequently expectorated, serving as an adhesive to attach the puparial case to a solid substrate during metamorphosis. Although this was first described almost 70 years ago, a detailed description of the morphology and mechanical properties of the glue is largely missing. Its main known physical property is that it is released as a watery liquid that quickly hardens into a solid cement. Here, we provide a detailed morphological and topological analysis of the solidified glue. We demonstrated that it forms a distinctive enamel-like plaque that is composed of a central fingerprint surrounded by a cascade of laterally layered terraces. The solidifying glue rapidly produces crystals of KCl on these alluvial-like terraces. Since the properties of the glue affect the adhesion of the puparium to its substrate, and so can influence the success of metamorphosis, we evaluated over 80 different materials for their ability to adhere to the glue to determine which properties favor strong adhesion. We found that the alkaline Sgs-glue adheres strongly to wettable and positively charged surfaces but not to neutral or negatively charged and hydrophobic surfaces. Puparia formed on unfavored materials can be removed easily without leaving fingerprints or cascading terraces. For successful adhesion of the Sgs-glue, the material surface must display a specific type of triboelectric charge. Interestingly, the expectorated glue can move upwards against gravity on the surface of freshly formed puparia via specific, unique and novel anatomical structures present in the puparial's lateral abdominal segments that we have named bidentia.

From Anomaly Detection to Defect Classification.

This paper proposes a new approach to defect detection system design focused on exact damaged areas demonstrated through visual data containing gear wheel images. The main advantage of the system is the capability to detect a wide range of patterns of defects occurring in datasets. The methodology is built on three processes that combine different approaches from unsupervised and supervised methods. The first step is a search for anomalies, which is performed by defining the correct areas on the controlled object by using the autoencoder approach. As a result, the differences between the original and autoencoder-generated images are obtained. These are divided into clusters using the clustering method (DBSCAN). Based on the clusters, the regions of interest are subsequently defined and classified using the pre-trained Xception network classifier. The main result is a system capable of focusing on exact defect areas using the sequence of unsupervised learning (autoencoder)-unsupervised learning (clustering)-supervised learning (classification) methods (U2S-CNN). The outcome with tested samples was 177 detected regions and 205 occurring damaged areas. There were 108 regions detected correctly, and 69 regions were labeled incorrectly. This paper describes a proof of concept for defect detection by highlighting exact defect areas. It can be thus an alternative to using detectors such as YOLO methods, reconstructors, autoencoders, transformers, etc.