Search Results

Cameras and image processing hardware are rapidly evolving technologies, which enable real-time applications for passenger cars, ground robots, and aerial vehicles. Visual odometry (VO) algorithms estimate vehicle position and orientation changes from the moving camera images. For ground vehicles, such as cars, indoor robots, and planetary rovers, VO can augment movement estimation from rotary wheel encoders. Feature-based VO relies on detecting feature points, such as corners or edges, in image frames as the vehicle moves. These points are tracked over frames and, as a group, estimate motion. Not all detected points are tracked since not all are found in the next frame. Even tracked features may not be correct since a feature point may map to an incorrect nearby feature point. This can depend on the driving scenario, which can include driving at high speed or in the rain or snow.

The need of weight reduction for fuel reduction and CO₂ regulations enforces the use of light-weight materials for structural parts also. The importance of reinforced composites will grow in this area. While the structural behavior and the simulation up to high strain-rate processes for those materials have been in the focus of investigation for many years, nowadays the simulation of high cycle fatigue behavior is getting important as well. Efficient fatigue analysis for metals was developed by understanding the microscopic behavior (crack nucleation and initiation) and bringing it to the macroscopic level by combining it with the matching test data (SN curves, etc.). Similar approaches can be applied to composite materials as well.

The Hybrid Cellular Automaton (HCA) algorithm is a generative design approach used to synthesize conceptual designs of crashworthy vehicle structures with a target mass. Given the target mass, the HCA algorithm generates a structure with a specific acceleration-displacement profile. The extended HCA (xHCA) algorithm is a generalization of the HCA algorithm that allows to tailor the crash response of the vehicle structure. Given a target mass, the xHCA algorithm has the ability to generate structures with different acceleration-displacement profiles and target a desired crash response. In order to accomplish this task, the xHCA algorithm includes two main components: a set of meta-parameters (in addition target mass) and surrogate model technique that finds the optimal meta-parameter values. This work demonstrates the capabilities of the xHCA algorithm tailoring acceleration and intrusion through the use of one meta-parameter (design time) and the use of Kriging-assisted optimization.

Simultaneous Localization and Mapping (SLAM) algorithms are extensively utilized within the field of autonomous navigation. In particular, numerous open-source Robot Operating System (ROS) based SLAM solutions, such as Gmapping, Hector, Cartographer etc., have simplified deployments in application. However, establishing the accuracy and precision of these ‘out-of-the-box’ SLAM algorithms is necessary for improving the accuracy and precision of further applications such as planning, navigation, controls. Existing benchmarking literature largely focused on validating SLAM algorithms based upon the quality of the generated maps. In this paper, however, we focus on examining the localization accuracy of existing 2-dimensional LiDAR based indoor SLAM algorithms. The fidelity of these implementations is compared against the OptiTrack motion capture system which is capable of tracking moving objects at sub-millimeter level precision.

Residual stress prediction arising from manufacturing processes provides paramount information for the fatigue performance assessment of components subjected to cyclic loading. The determination of the material model to be applied in the numerical model should be taken carefully. This study focuses on the estimation of residual stresses generated after deep rolling of cast iron crankshafts. The researched literature on the field employs the available commercial material codes without closer consideration on their reverse loading capacities. To mitigate this gap, a single element model was used to compare potential material models with tensile-compression experiments. The best fit model was then applied to a previously developed crankshaft deep rolling numerical model. In order to confront the simulation outcomes, residual stresses were measured in two directions on real crankshaft specimens that passed through the same modeled deep rolling process.

In order to explore approaches for improving object detection accuracy in intelligent vehicle system, we exploit super-resolution techniques. A novel method is proposed to confirm the conjecture whether some popular super-resolution networks used for environmental perception of intelligent vehicles and robots can indeed improve the detection accuracy. COCO dataset which contains images from complex ordinary environment is utilized for the verification experiment, due to it can adequately verify the generalization of each algorithm and the consistency of experimental results. Using two representative object detection networks to produce the detection results, namely Faster R-CNN and YOLOv3, we devise to reduce the impact of resizing operation. The two networks allow us to compare the performance of object detection between using original and super-resolved images. We quantify the effect of each super-resolution techniques as well.

Spot Welds are a category of welds used extensively in automotive structures, normally for metals. The fatigue analysis of such spot welds can be evaluated using (a) the Point 2 Point (P2P) method where a beam or bar is used to connect the 2 surfaces being joined, (b) a more modern approach where the 1D element is replaced with an “equivalent” brick element, or (c) a third approach that falls somewhere between where a “spider” and circular ring of elements, is used to represent the spot weld. In all 3 cases there is an assumption that the cross section is circular. For some specialist cases such as plastic connectors, the cross section is not circular so a new user defined weld is proposed. This paper will describe the approach that is based on the concept that a user generated tensor line can be used (equivalent to the theoretical Force/Moment to stress algorithms built into the P2P approach) along with special S-N curves create for different joint shapes.

Collaborative robots are more and more used in smart manufacturing because of their capability to work beside and collaborate with human workers. With the deployment of these robots, manufacturing tasks are more inclined to be accomplished by multiple humans and multiple robots (MH-MR) through teaming effort. In such MH-MR collaboration scenarios, the task distribution among the multiple humans and multiple robots is very critical to efficiency. It is also more challenging due to the heterogeneity of different agents. Existing approaches in task distribution among multiple agents mostly consider humans with assumed or known capabilities. However human capabilities are always changing due to various factors, which may lead to suboptimal efficiency. Although some researches have studied several human factors in manufacturing and applied them to adjust the robot task and behaviors.

In the high frequency range, the SEA method has been applied to air borne path with success to predict both internal and external sound environment. Nevertheless, structure-borne prediction is still at issue -especially for cars, in the range 200 to 2000 Hz- as results are widely dependant on subsystem partition and validity of various assumptions required by SEA. Experimental SEA test technique (ESEA), applied to car bodies, has brought to the fore that SEA power balanced equations could robustly describe structure-borne noise. To make ESEA predictive, the database of measured FRF is simply replaced and enlarged by synthesized data generated from a finite element (FE) model and a selected observation grid of nodes. This technique, called Virtual SEA (VSEA), has been presented at SAE/NVC 2003.

The paper discusses two methods to implement error compensation framework for NC machining. In the first case a novel real-time co-interpolator is utilized that has been developed and demonstrated on a machine tool equipped with an open architecture controller (OAC). The second solution features near-time reprocessing of the NC programs, that is more suitable for existing machining systems. A P-type, iterative learning control (ILC) algorithm is also presented for calculating the error compensation values. The paper concludes with the results of machining tests, showing the effectiveness of the error compensation methodologies.

Due to alternate local heating and cooling (Thermal cycles) during welding, complex stresses occur and so residual stresses and angular distortion occur after welding. The residual stresses and angular distortion present in the welded components cause internal cracks and mismatching of welded automobile structures. Also the tensile residual stresses in area near the weldment may cause fractures under certain conditions whereas compressive residual stresses in the base plate may reduce buckling strength of structural members. Hence a welding system which is capable of predicting and acting according to it is the need of the hour. It is difficult to obtain a more reliable complete analytical solution to predict angular distortion and maximum residual stresses for a wide range of welding processes, materials and process parameters.

A math-based approach is now in use to reduce transmission control algorithm development time and prototyping iterations, and to improve transmission control algorithm and software quality. This paper will describe the application of this efficient approach to the development of a transmission control algorithm for production. The processes for algorithm design, closed-loop and rapid prototype testing, auto-code generation of production software, and potential hardware-in-the-loop verification are described with references to supporting development tools. This approach enables upfront, math-based development and performance assessment of transmission control algorithms and their interactions with other powertrain and vehicle control algorithms. Examples will illustrate the efficiency of the control algorithm development approach over a wide range of conditions.

This paper is a case study of a successful application of a particular type of machine vision system to an aerospace manufacturing operation. The task for the vision system is to find holes very accurately. The system performs this task using a novel combination of laser triangulation and grayscale area analysis. Using a custom sensor, incorporating a high resolution digital camera, structured and unstructured illumination, together with advanced image processing algorithms has resulted in a measurement system which has excellent accuracy and repeatability.

A method to predict gap distribution, can deformation and mounting force of catalytic converter during assembling and operation cycles has been developed using ABAQUS contact algorithm with user subroutine for material properties. Inherent in the methodology is the constitutive model for both vermiculite mat and wire mesh mounting materials, which is able to describe their nonlinear and thermal behaviors and shows good agreement with test results. A design optimization procedure is presented to achieve uniform gap design of can and substrate. The technology will enable engineers to generate robust converter can designs, substrate shape and stamping tools for minimum manufacturing failure rate and maximum durability performance once a mounting material is selected.

Workstation design is one of the most essential components of proactive ergonomics, and digital human models have gained increasing popularity in the analysis and design of current and future workstations (Chaffin 2001). Using digital human technology, it is possible to simulate interactions between humans and current or planned workstations, and conduct quantitative ergonomic analyses based on realistic human postures and motions. Motion capture has served as the primary means by which to acquire and visualize human motions in a digital environment. However, motion capture only provides motions for a specific person performing specific tasks. Albeit useful, at best this allows for the analysis of current or mocked-up workstations only. The ability to subsequently modify these motions is required to efficiently evaluate alternative design possibilities and thus improve design layouts.

Machining fixtures are used to locate and constrain a workpiece during a machining operation. To ensure that the workpiece is manufactured according to specified dimensions and tolerances, it must be appropriately located and clamped. Minimizing workpiece and fixture tooling deflections due to clamping and cutting forces in machining is critical to the machining accuracy. An ideal fixture design maximizes locating accuracy and workpiece stability, while minimizing displacements. The purpose of this research is to develop a method for modeling workpiece boundary conditions and applied loads during a machining process, analyze modular fixture tool contact area deformation and optimize support locations, using finite element analysis (FEA). The workpiece boundary conditions are defined by locators and clamps. The locators are placed in a 3-2-1 fixture configuration, constraining all degrees of freedom of the workpiece and are modeled using linear spring-gap elements.

The system identification procedure is a powerful and flexible tool for the modeling of dynamic systems. This paper implements the theory of parametric identification in order to estimate a valid model of a flexible robotic arm. For this purpose experimental data is used for the estimation of ARMAX SISO models. A two-stages identification procedure (non-parametric & parametric) provides an insight about the system under identification. In the first stage, known signal analysis methods are applied (correlation-spectral analysis) for the estimation of frequencies and frequency response, and in the second stage, the estimation of ARMAX models is performed in order to fit a transfer function model to collected input-output data set. For the estimation of model's coefficients, use of Evolutionary Algorithms is implemented.

This paper describes and demonstrates the use of an assembly centric design algorithm as an aid to achieving minimal hard tooling assembly concepts. The algorithm consists of a number of logically ordered design methodologies and also aids the identification of other enabling technologies. Included in the methodologies is an innovative systems analysis tool that enables the comparison of alternative assembly concepts, and the prediction and control of the total assembly error, at the outline stage of the design.

A current general need within Six Sigma methodologies is to utilize statistical methods including experimental design in the confirmation of new processes and their parameters. This is typically done in the improve step of the DMAIC process. This need is even more evident in microjoining (small scale resistance welding) due to the number and complexity of the process variables. This paper outlines the improve step of a Six Sigma project in which statistical methods are applied to a microjoining process. These statistical methods include linear experimental design, regression analysis with linear transformation and mathematical modeling. The paper documents the methodology used to establish process parameters in microjoining of an electrical lead frame design.

We describe an optimization model developed by Ford Motor Company to reallocate stamped parts between facilities when business conditions change. How can the business meet new targets when demand starts to exceed existing capacity? Likewise, how can it respond when demand is lower than expected? Sometimes the business can reduce costs by transferring production to a different location or by outsourcing parts. We describe in this paper how mathematical optimization can identify solutions that balance both logistical and outsourcing costs. We explain the algorithm and demonstrate with a small example how it recommends sourcing plans that minimize cost.