Abstract
Diabetes, a chronic condition affecting millions of people, requires ongoing medical care and treatment, which can place a significant financial burden on society, directly and indirectly. In this paper we propose a vision-robotics system for the automatic assessment of the diabetic foot, one the exams used for the disease management. We present and discuss various computer vision techniques that can support the core operation of the system. U-Net and Segnet, two popular convolutional network architectures for image segmentation are applied in the current case. Hardcoded and machine learning pipelines are explained and compared using different metrics and scenarios. The obtained results show the advantages of the machine learning approach but also point to the importance of hard coded rules, especially when well know areas, such as the human foot, are the systems’ target. Overall, the system achieved very good results, paving the way to a fully automated clinical system. © 2024, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Abstract
Logistics chains are being increasingly developed due to several factors, among which the exponential growth of e-commerce. Crossdocking is a logistics strategy used by several companies from varied economic sectors, applied in warehouses and distribution centres. In this context, it is the objective of the “CrossLog – Automatic Mixed-Palletizing for Crossdocking Logistics Centers” Project, to investigate and study an automated and collaborative crossdocking system, capable of moving and managing the flow of products within the warehouse in the fastest and safest way. In its scope, this paper describes the concept and architecture envisioned for the crossdocking system developed in the scope of the CrossLog Project. One of its main distinguishing characteristics is the use of Autonomous Mobile Robots for performing much of the operations traditionally performed by human operators in today’s logistics centres. © 2024, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Abstract
Due to the lack of unskilled labour force that has been verified in the last years, several processes have been automated, both at industrial and services level. In terms of logistics tasks and transport of materials, it is increasingly common to use mobile robots, given the advantages that this equipment presents. This is also the case in airports, where the adoption of these vehicles to perform several tasks is becoming visible. Considering the possibility of using mobile robots to transport luggage at the Francisco Sa, Carneiro Airport, this paper presents the development of a simulation model and the analysis of several scenarios, with different number of vehicles, in order to understand the time that passengers would have to wait for their luggage, in case this task is automated. The final objective is to determine the number of vehicles required and the changes that need to be made to the airport's operation in order to ensure a level of service identical to (or better than) that currently achieved, with these operations being carried out by human operators.

Abstract
Logistics chains are being increasingly developed due to several factors, among which the exponential growth of e-commerce. Cross-docking is a logistics strategy used by several companies from varied economic sectors, applied in warehouses and distribution centres. In this context, it is the objective of the CrossLog - Automatic Mixed-Palletizing for Crossdocking Logistics Centers Project, to investigate and study an automated and collaborative crossdocking system, capable of moving and managing the flow of products within the warehouse in the fastest and safest way. In its scope, this paper describes the modelling and simulation of a crossdocking system, with an integrated AS/RS, to analyse possible alternatives including not only the fast movement of products, but also its storage in case needed. Different scenarios were modelled and simulated, on the FlexSim software, and the obtained results for each one were critically analysed to draw conclusions on the best storage policy.

Abstract
Bin picking based on deep learning techniques is a promising approach that can solve several analytical methods problems. These systems can provide accurate solutions to bin picking in cluttered environments, where the scenario is always changing. This article proposes a robust and accurate system for segmenting bin picking objects, employing an easy configuration procedure to adjust the framework according to a specific object. The framework is implemented in Robot Operating System (ROS) and is divided into a detection and segmentation system. The detection system employs Mask R-CNN instance neural network to identify several objects from two dimensions (2D) grayscale images. The segmentation system relies on the point cloud library (PCL), manipulating 3D point cloud data according to the detection results to select particular points of the original point cloud, generating a partial point cloud result. Furthermore, to complete the bin picking system a pose estimation approach based on matching algorithms is employed, such as Iterative Closest Point (ICP). The system was evaluated for two types of objects, knee tube, and triangular wall support, in cluttered environments. It displayed an average precision of 79% for both models, an average recall of 92%, and an average IOU of 89%. As exhibited throughout the article, this system demonstrates high accuracy in cluttered environments with several occlusions for different types of objects.