Roman Mykhailyshyn

Modern neural networks have made great strides in recognising objects in images and are widely used in defect detection. However, the output of a neural network strongly depends on both the training dataset and the conditions under which the image was acquired for analysis. We have developed a software–hardware method for evaluating the effect of variable lighting on the results of defect recognition using a neural network model. The proposed approach allows us to analyse the recognition results of an existing neural network model and identify the optimal range of illumination at which the desired defects are recognised most consistently. For this purpose, we analysed the variability in quantitative parameters (area and orientation) of damage obtained at different degrees of illumination for two different light sources: LED and conventional incandescent lamps. We calculated each image’s average illuminance and quantitative parameters of recognised defects. Each set of parameters represents the results of defect recognition for a particular illuminance level of a given light source. The proposed approach allows the results obtained using different light sources and illumination levels to be compared and the optimal source type/illuminance level to be figured out. This makes implementing a defect detection environment that allows the best recognition accuracy and the most controlled product quality possible. An analysis of a steel sheet surface showed that the best recognition result was achieved at an illuminance of ~200 lx. An illuminance of less than ~150 lx does not allow most defects to be recognised, whereas an illuminance larger than ~250 lx increases the number of small objects that are falsely recognised as defects.

The field of continuum robotics is rapidly developing. The development of new kinematic structures, locomotion principles and control strategies is driving the development of new types of sensors and sensing methodologies. The sensing in continuum robots can be divided into shape perception and environment perception. The environment perception is focusing on sensing the interactions between the robot and environment. These sensors are often embedded on an outer layer of the robots, so the interactions can be detected. The shape perception is sensing the robot’s shape using various principles. There are three main groups of sensors that use the properties of electricity, magnetism and optics to measure the shape of the continuum robots. The sensors based on measuring the properties of electricity are often based on measuring the electrical resistance or capacitance of the flexible sensor. Sensors based on magnetism use properties of permanent magnets or coils that are attached to the robot. Their magnetic field, flux or other properties are then tracked, and shape reconstruction can be performed. The last group of sensors is mostly based on leveraging the properties of traveling light through optical fibers. There are multiple objectives of this work. Objective number one is to clearly categorize the sensors and make a clear distinction between them. Objective number two is to determine the trend and progress of the sensors used in continuum robotics. And finally, the third objective is to define the challenges that the researchers are currently facing. The challenges of sensing the shape or the interaction with the environment of continuum robots are currently in the miniaturization of existing sensors and the development of novel sensing methods.

This article is focused on the problematics of path planning, which means finding the optimal path between two points in a known environment with obstacles. The proposed path-planning method uses the wavefront algorithm, and two modifications are implemented and verified. The first modification is the removal of redundant waypoints. The first modification is applied because the wavefront algorithm generates redundant waypoints. These waypoints cause unnecessary changes in the direction of movement. The second one is smoothing the generated trajectory using B-spline curves. The reason for applying the second modification is that trajectory generated by the wavefront algorithm is in the form of the polyline, which is inadequate in terms of the smoothness of the robot’s motion. The verification of the proposed method is performed in environments with different densities of obstacles compared with standard Dijkstra’s and A* algorithms.