Search results

Workpiece sorting is a key link in industrial production lines. The vision-based workpiece sorting system is non-contact and widely applicable. The detection and recognition of workpieces are the key technologies of the workpiece sorting system. To introduce deep learning algorithms into workpiece detection and improve detection accuracy, this paper aims to propose a workpiece detection algorithm based on the single-shot multi-box detector (SSD).

Propose a multi-feature fused SSD network for fast workpiece detection. First, the multi-view CAD rendering images of the workpiece are used as deep learning data sets. Second, the visual geometry group network was trained for workpiece recognition to identify the category of the workpiece. Third, this study designs a multi-level feature fusion method to improve the detection accuracy of SSD (especially for small objects); specifically, a feature fusion module is added, which uses “element-wise sum” and “concatenation operation” to combine the information of shallow features and deep features.

Experimental results show that the actual workpiece detection accuracy of the method can reach 96% and the speed can reach 41 frames per second. Compared with the original SSD, the method improves the accuracy by 7% and improves the detection performance of small objects.

This paper innovatively introduces the SSD detection algorithm into workpiece detection in industrial scenarios and improves it. A feature fusion module has been added to combine the information of shallow features and deep features. The multi-feature fused SSD network proves the feasibility and practicality of introducing deep learning algorithms into workpiece sorting.

To improve production efficiency, industrial robots are expected to replace humans to complete the traditional manual operation on grasping, sorting and assembling workpieces. These implementations are closely related to the accuracy of workpiece location. However, workpiece location methods based on conventional machine vision are sensitive to the factors such as light intensity and surface roughness. To enhance the robustness of the workpiece location method and improve the location accuracy, a workpiece location algorithm based on improved Single Shot MultiBox Detector (SSD) is proposed.

The proposed algorithm integrates a weighted bi-directional feature pyramid network into SSD. A feature fusion architecture is structured by the combination of low-resolution, strong semantic features and high-resolution, weak semantic features. The architecture is built through a top-down pathway, bottom-up pathway, lateral connections and skip connections. To avoid treating all features equally, learnable weights are introduced into each feature layer to characterize its importance. More detailed information from the low-level layers is injected into the high-level layers, which could improve the accuracy of workpiece location.

It is found that the maximum location error at the center point calculated from the proposed algorithm is decreased by more than 22% compared with that of the SSD algorithm. Besides, the average location error evolves a decrease by at least 5%. In the trajectory prediction experiment of the workpiece center point, the results of the proposed algorithm demonstrate that the average location error is below 0.13 mm and the maximum error is no more than 0.23 mm.

In this work, a workpiece location algorithm based on improved SSD is developed to extract the center point of the workpiece. The results demonstrate that the proposed algorithm is beneficial for workpiece location. The proposed algorithm can be readily used in a variety of workpieces or adapted to other similar tasks.

This study aims to solve the problem of job scheduling and multi automated guided vehicle (AGV) cooperation in intelligent manufacturing workshops.

In this study, an algorithm for job scheduling and cooperative work of multiple AGVs is designed. In the first part, with the goal of minimizing the total processing time and the total power consumption, the niche multi-objective evolutionary algorithm is used to determine the processing task arrangement on different machines. In the second part, AGV is called to transport workpieces, and an improved ant colony algorithm is used to generate the initial path of AGV. In the third part, to avoid path conflicts between running AGVs, the authors propose a simple priority-based waiting strategy to avoid collisions.

The experiment shows that the solution can effectively deal with job scheduling and multiple AGV operation problems in the workshop.

In this paper, a collaborative work algorithm is proposed, which combines the job scheduling and AGV running problem to make the research results adapt to the real job environment in the workshop.

This paper aims to propose an identification method based on monocular vision for cylindrical parts in cluttered scene, which solves the issue that iterative closest point (ICP) algorithm fails to obtain global optimal solution, as the deviation from scene point cloud to target CAD model is huge in nature.

The images of the parts are captured at three locations by a camera amounted on a robotic end effector to reconstruct initial scene point cloud. Color signatures of histogram of orientations (C-SHOT) local feature descriptors are extracted from the model and scene point cloud. Random sample consensus (RANSAC) algorithm is used to perform the first initial matching of point sets. Then, the second initial matching is conducted by proposed remote closest point (RCP) algorithm to make the model get close to the scene point cloud. Levenberg Marquardt (LM)-ICP is used to complete fine registration to obtain accurate pose estimation.

The experimental results in bolt-cluttered scene demonstrate that the accuracy of pose estimation obtained by the proposed method is higher than that obtained by two other methods. The position error is less than 0.92 mm and the orientation error is less than 0.86°. The average recognition rate is 96.67 per cent and the identification time of the single bolt does not exceed 3.5 s.

The presented approach can be applied or integrated into automatic sorting production lines in the factories.

The proposed method improves the efficiency and accuracy of the identification and classification of cylindrical parts using a robotic arm.

Application of basic processing equipment, providing for high‐levelled combination and concentration of operations, together with automatic program‐controlled manipulators of various types offers possibilities for creating automated production systems which differ in purpose, structure, complexity and a degree of automation of the main and auxiliary operations and control as well.

SUITABILITY of materials and their usage can be a profitable source of investigation for the methods engineer. And an important function too, for it implies not only an examination of the most suitable material in relation to the product to be processed, but also, methods of storage, inspection, and the economic utilisation of that material at all stages. Any of these aspects involves a major investigation if the firm is interested in discovering untapped sources of economy. It is one of the avenues not often explored by work study, but there is no valid reason why it should remain a sort of no‐man's‐land.

Fixture layout design is concerned with immobilization of the workpiece (engine mount bracket) during machining such that the workpiece elastic deformation is reduced. The fixture holds the workpiece through the positioning of fixturing elements that causes the workpiece elastic deformation, in turn, leads to the form and dimensional errors and increased machining cost. The fixture layout has the major impact on the machining accuracy and is the function of the fixturing position. The position of the fixturing elements, key aspects, needed to be optimized to reduce the workpiece elastic deformation. The purpose of this study is to evaluate the optimized fixture layout for the machining of the engine mount bracket.

In this research work, using the finite element method (FEM), a model is developed in the MATLAB for the fixture-workpiece system so that the workpiece elastic deformation is determined. The artificial neural network (ANN) is used to develop an empirical model. The results of deformation obtained for different fixture layouts from FEM are used to train the ANN and finally the empirical model is developed. The model capable of predicting the deformation is embedded to the evolutionary optimization techniques, capable of finding local and global optima, to optimize the fixture layouts and to find the robust one.

For efficient optimization of the fixture layout parameters to obtain the least possible deformation, ant colony algorithm (ACA) and artificial bee colony algorithm (ABCA) are used and the results of deformation obtained from both the optimization techniques are compared for the best results.

A MATLAB-based FEM technique is able to provide solutions when the repeated modeling and simulations required i.e. modeling of fixture layouts (500 layouts) for every variation in the parameters requires individual modeling and simulation for the output requirement in any FEM-based software’s (ANSYS, ABACUS). This difficulty is reduced in this research. So that the MATLAB-based FEM modeling, simulation and optimization is carried out to determine the solutions for the optimized fixture layout to reach least deformation.

Many a time the practicability of the machining/mechanical operations are difficult to perform costly and time-consuming when more number of experimentations are required. To sort out the difficulties the computer-based automated solution techniques are highly required. Such kind of research over this study is presented for the readers.

A MATLAB-based FEM modeling and simulation technique is used to obtain the fixture layout optimization. ANN-based empirical model is developed for the fixture layout deformation that creates a hypothesis for the fixture layout system. ACA and ABCA are used for optimizing the fixture layout parameters and are compared for the best algorithm suited for the fixture layout system.

A finite‐element model for calculating the die temperature profile for a hot‐forging operation is presented. The workpiece is modelled as a thermo‐viscoplastic material, while the dies are considered undeformable. Heat transfer between the dies and the workpiece is modelled using an iteratively coupled, fixed‐point calculation of the temperature in each domain. Transfer of temperature boundary conditions across contact interfaces is performed for non‐coincident meshes, using a boundary integration point contact analysis. Two industrial‐type examples are presented. In the first example, the effectiveness of the transfer of the temperature boundary conditions for a non steady‐state forging process is evaluated and determined to be satisfactory. Then weakly‐ and strongly‐coupled temperature resolutions are compared. It was found that the strongly‐coupled resolution may be necessary in order to obtain reasonably accurate results. In the second example, the weakly‐coupled resolution is compared to a constant‐temperature die approach for a relatively slow forging process, which shows the influence of the die temperature on the flow of the material.

Four assembly and four manual workstations make up a line for producing four types of potentiometer.

The purpose of this paper is to present a robotic off-line programming method for freeform surface grinding based on visualization toolkit (VTK). Nowadays, manual grinding and traditional robot on-line programming are difficult to ensure the surface grinding accuracy, thus off-line programming is gradually used in grinding, however, several problems are needed to be resolved which include: off-programming environment depends on the third-party CAD software, leads to insufficient self-development flexibility; single support for robot type or workpiece model format contributes to lack of versatility; grinding point data depends on external data calculation and import process, causes human-computer interaction deterioration.

In this method, the visualization pipeline and observer/command mode of VTK are used to display the 3D model of the robot grinding system and pick up the workpiece surfaces to be grinded respectively. Two groups of cutter planes with equidistant spacing are created to form the grinding nodes on the surface, and the extraction method for the position and posture of the nodes is proposed. Furthermore, the position and posture of discretized points along the grinding curve are obtained by B-spline curve interpolation and quaternion spherical linear interpolation respectively. Finally, the motion simulation is realized by robot inverse kinematics.

Through a watch case grinding experiment, the results show that the proposed method based on VTK can achieving high precision grinding effect, which is obviously better than traditional method.

The proposed method is universal which does not depend on the specific forms of surface, and all calculations in simulation are completed within the system, avoiding tedious external data calculation and import process. The grinding trajectory can be generated only by the mouse picking operation without relying on the other third-party CAD software.