Abstract
Machine vision systems are widely used in assembly lines for providing sensing abilities to robots to allow them to handle dynamic environments. This paper presents a comparison of 3D sensors for evaluating which one is best suited for usage in a machine vision system for robotic fastening operations within an automotive assembly line. The perception system is necessary for taking into account the position uncertainty that arises from the vehicles being transported in an aerial conveyor. Three sensors with different working principles were compared, namely laser triangulation (SICK TriSpector1030), structured light with sequential stripe patterns (Photoneo PhoXi S) and structured light with infrared speckle pattern (Asus Xtion Pro Live). The accuracy of the sensors was measured by computing the root mean square error (RMSE) of the point cloud registrations between their scans and two types of reference point clouds, namely, CAD files and 3D sensor scans. Overall, the RMSE was lower when using sensor scans, with the SICK TriSpector1030 achieving the best results (0.25 mm ± 0.03 mm), the Photoneo PhoXi S having the intermediate performance (0.49 mm ± 0.14 mm) and the Asus Xtion Pro Live obtaining the higher RMSE (1.01 mm ± 0.11 mm). Considering the use case requirements, the final machine vision system relied on the SICK TriSpector1030 sensor and was integrated with a collaborative robot, which was successfully deployed in an vehicle assembly line, achieving 94% success in 53,400 screwing operations.

Abstract
The Multi-View Stereo (MVS) is a key process in the photogrammetry workflow. It is responsible for taking the camera's views and finding the maximum number of matches between the images yielding a dense point cloud of the observed scene. Since this process is based on the matching between images it greatly depends on the ability of features matching throughout different images. To improve the matching performance several researchers have proposed the use of Convolutional Neural Networks (CNNs) to solve the MVS problem. Despite the progress in the MVS problem with the usage of CNNs, the Video RAM (VRAM) consumption within these approaches is usually far greater than classical methods, that rely more on RAM, which is cheaper to expand than VRAM. This work then follows the progress made in CasMVSNet in the reduction of GPU memory usage, and further study the changes in the feature extraction process. The Average Group-wise Correlation is used in the cost volume generation, to reduce the number of channels in the cost volume, yielding a reduction in GPU memory usage without noticeable penalties in the result. The deformable convolutions are applied in the feature extraction network to augment the spatial sampling locations with learning offsets, without additional supervision, to further improve the network's ability to model transformations. The impact of these changes is measured using quantitative and qualitative tests using the DTU and the Tanks and Temples datasets. The modifications reduced the GPU memory usage by 32% and improved the completeness by 9% with a penalty of 6.6% in accuracy on the DTU dataset. © 2022

Abstract

Abstract
With the continuously growing usage of collaborative robots in industry, the need for achieving a seamless human-robot interaction has also increased, considering that it is a key factor towards reaching a more flexible, effective, and efficient production line. As a prominent and prospective tool to support the human operator to understand and interact with robots, Augmented Reality (AR) has been employed in numerous human-robot collaborative and cooperative industrial applications. Therefore, this systematic literature review critically appraises 32 papers' published between 2016 and 2021 to identify the main employed AR technologies, outline the current state of the art of augmented reality for human-robot collaboration and cooperation, and point out future developments for this research field. Results suggest that this is still an expanding research field, especially with the advent of recent advancements regarding head-mounted displays (HMDs). Moreover, projector-based and HMDs developed approaches are showing promising positive influences over operator-related aspects such as performance, task awareness, and safety feeling, even though HMDs need further maturation in ergonomic aspects. Further research should focus on large-scale assessment of the proposed solutions in industrial environments, involving the solution's target audience, and on establishing standards and guidelines for developing AR assistance systems.

Abstract
Odometry calibration adjusts the kinematic parameters or directly the robot's model to improve the wheeled odometry accuracy. The existent literature considers in the calibration procedure only one steering geometry (differential drive, Ackerman/tricycle, or omnidirectional). Our method, the OptiOdom calibration algorithm, generalizes the odometry calibration problem. It is developed an optimization-based approach that uses the improved Resilient Propagation without weight-backtracking (iRprop-) for estimating the kinematic parameters using only the position data of the robot. Even though a calibration path is suggested to be used in the calibration procedure, the OptiOdom method is not path-specific. In the experiments performed, the OptiOdom was tested using four different robots on a square, arbitrary, and suggested calibration paths. The OptiTrack motion capture system was used as a ground-truth. Overall, the use of OptiOdom led to improvements in the odometry accuracy (in terms of maximum distance and absolute orientation errors over the path) over the existent literature while being a generalized approach to the odometry calibration problem. The OptiOdom and the methods from the literature implemented in the article are available in GitHub as an open-source repository.