Identifying High-Value CAD Models: An Exploratory Study on Dimensional Variability As Complexity Indicator

Digital product data quality has proven to be a unifying theme in designing and reusing efficient products, particularly in the context of the Model-Based Enterprise (MBE). More specifically, the quality of the master model (usually a history-based parametric model) is critical, as it determines the quality of all secondary models used in subsequent downstream processes. However, no quantitative metrics exist that can provide a reliable assessment of quality at a high semantic level. In this paper, we introduce dimensional variability as a quality indicator for parametric models that connects the effective variability range of the dimensional constraints in a model to the robustness and flexibility of the parametric geometry, which determines its reusability. As a validation effort, we report the results of a study where a set of parametric models of varying complexity was analyzed, and discuss the significance of the links between the proposed metric and various aspects of the internal graph structure of the CAD model.

Brain Computer Interface Robotic Co-Workers: Defective Part Picking System

Industrial robotic co-workers are robots that can work with human being in an unstructured environment. Such robots, must be able to assist human operators in a seamless way without receiving specific instructions. Robotic co-workers can open entirely new application fields in manufacturing as demonstrated in this paper. We designed such an industrial co-robot to pick up defective parts by simply monitoring a human operator directly through a brain computer interface (BCI). By constantly monitoring the operator using BCI sensors, the robotic co-worker can sense when an operator notices a defective part and then moves to remove the part from a moving conveyor with no direct instruction from the operator. The robot, equipped with an RGB camera, recognizes the part, tracks the position and generates accurate motion plan. We demonstrated the system using a human subject study.

Co-Design of Supply Chain Network and Subassembly Planning Considering the Reconfiguration of Supply Chain Structure for Factory-in-a-Box Manufacturing

As the market demand becomes more diversified and dynamic, the requirements for manufacturing systems feature a high degree of flexibility, low cost, low volume, and short delivery times. One emerging way for such flexible manufacturing is so-called “factory-in-a-box,” by which production modules are installed in a container and transported by a vehicle. The factory-in-a-box manufacturing poses a unique challenge to manufacturing supply chain network since the ease of supply chain reconfiguration when the vehicle moves to a different production site has become a major concern in addition to transportation cost and delivery time. The supply chain design is further complicated by the fact that it is coupled with subassembly planning in manufacturing, which determines appropriate subassembly modules assigned to suppliers. As such, it is critical to understand the interaction between supply chain network reconfigurability and subassembly planning. This paper develops a model using a set of decision variables to jointly characterize the topology of supply chain network and subassembly planning. A binary nonlinear programming model has been developed for the concurrent optimization of subassembly planning and supply chain network with the consideration of reconfiguration of the supply chain structure. One numerical case study was conducted to demonstrate the proposed model by providing a quantitative guideline of reconfiguring supply chain network when the final production site (on a vehicle) changes locations.

Monitoring, Diagnostics, and Prognostics for Robot Tool Center Accuracy Degradation

Over time, robots degrade because of age and wear, leading to decreased reliability and increasing potential for faults and failures; this negatively impacts robot availability. Economic factors motivate facilities and factories to improve maintenance operations to monitor robot degradation and detect faults and failures, especially to eliminate unexpected shutdowns. Since robot systems are complex, with sub-systems and components, it is challenging to determine these constituent elements’ specific influence on the overall system performance. The development of monitoring, diagnostic, and prognostic technologies (collectively known as Prognostics and Health Management (PHM)), can aid manufacturers in maintaining the performance of robot systems by providing intelligence to enhance maintenance and control strategies. This paper presents the strategy of integrating top level and component level PHM to detect robot performance degradation (including robot tool center accuracy degradation), supported by the development of a four-layer sensing and analysis structure. The top level PHM can quickly detect robot tool center accuracy degradation through advanced sensing and test methods developed at the National Institute of Standards and Technology (NIST). The component level PHM supports deep data analysis for root cause diagnostics and prognostics. A reference data set is collected and analyzed using the integration of top level PHM and component level PHM to understand the influence of temperature, speed, and payload on robot’s accuracy degradation.

Towards Continuous Production of Shaped Honeycombs

Honeycomb sandwich panels are widely used for high performance parts subject to bending loads, but their manufacturing costs remain high. In particular, for parts with non-flat, non-uniform geometry, honeycombs must be machined or thermoformed with great care and expense. The ability to produce shaped honeycombs would allow sandwich panels to replace monolithic parts in a number of high performance, space-constrained applications, while also providing new areas of research for structural optimization, distributed sensing and actuation, and on-site production of infrastructure. Previous work has shown methods of directly producing shaped honeycombs by cutting and folding flat sheets of material. This research extends these methods by demonstrating work towards a continuous process for the cutting and folding steps of this process. An algorithm for producing a manufacturable cut-and-fold pattern from a three-dimensional volume is designed, and a machine for automatically performing the required cutting and parallel folding is proposed and prototyped. The accuracy of the creases placed by this machine is characterized and the impact of creasing order is demonstrated. Finally, a prototype part is produced and future work is sketched towards full process automation.

A Two-Layer Long Short-Term Memory Network for Bottleneck Prediction in Multi-Job Manufacturing Systems

Throughput bottlenecks define and constrain the productivity of a production line. Prediction of future bottlenecks provides a great support for decision-making on the factory floor, which can help to foresee and formulate appropriate actions before production to improve the system throughput in a cost-effective manner. Bottleneck prediction remains a challenging task in literature. The difficulty lies in the complex dynamics of manufacturing systems. There are multiple factors collaboratively affecting bottleneck conditions, such as machine performance, machine degradation, line structure, operator skill level, and product release schedules. These factors impact on one another in a nonlinear manner and exhibit long-term temporal dependencies. State-of-the-art research utilizes various assumptions to simplify the modeling by reducing the input dimensionality. As a result, those models cannot accurately reflect complex dynamics of the bottleneck in a manufacturing system. To tackle this problem, this paper will propose a systematic framework to design a two-layer Long Short-Term Memory (LSTM) network tailored to the dynamic bottleneck prediction problem in multi-job manufacturing systems. This neural network based approach takes advantage of historical high dimensional factory floor data to predict system bottlenecks dynamically considering the future production planning inputs. The model is demonstrated with data from an automotive underbody assembly line. The result shows that the proposed method can achieve higher prediction accuracy compared with current state-of-the-art approaches.

Automated Surface Inspection Using 3D Point Cloud Data in Manufacturing: A Case Study

With the latest advancements in three-dimensional (3D) measurement technologies, obtaining 3D point cloud data for inspection purposes in manufacturing is becoming more common. While 3D point cloud data allows for better inspection capabilities, their analysis is typically challenging. Especially with unstructured 3D point cloud data, containing coordinates at random locations, the challenges increase with higher levels of noise and larger volumes of data. Hence, the objective of this paper is to extend the previously developed Adaptive Generalized Likelihood Ratio (AGLR) approach to handle unstructured 3D point cloud data used for automated surface defect inspection in manufacturing. More specifically, the AGLR approach was implemented in a practical case study to inspect twenty-seven samples, each with a unique fault. These faults were designed to cover an array of possible faults having three different sizes, three different magnitudes, and located in three different locations. The results show that the AGLR approach can indeed differentiate between non-faulty and a varying range of faulty surfaces while being able to pinpoint the fault location. This work also serves as a validation for the previously developed AGLR approach in a practical scenario.

Design of a Flexure-Based Active Fixture System for Precision Robotic Deburring

This paper presents on the design, prototyping and testing of a flexure-based active workpiece fixture system for precision robotic deburring. Current industrial robotic manipulators suffer from poor positioning accuracy, which makes precision tasks such as deburring, polishing and grinding challenging. Together, the robotic manipulator and the proposed active work fixture will create a dual-stage positioning system for precision tasks where position/force control is crucial. The main application is robotic deburring, which demands positioning accuracy and high compliance over large cutting forces. This first prototype active fixture system is designed as a planar motion table that is supported by parallel flexures, driven by voice-coil actuators, and uses high-resolution laser displacement pickups facilitate accurate motion generation with great backdrivability for force control. The theory behind the proposed design is shown, and a prototype is then used to validate performance. Overall the prototype flexure stage achieves a total stroke of 1 mm and a bandwidth of 21 Hz.

Study on Improving Accuracy in Robotic Milling of Aluminum Alloy

Nowadays, the application of using industrial robots in manufacture is a diminutive due to its own low rigidity and low stiffness. This leads to high level of vibrations that limits the quality and the precision of the workpiece. So they are usually used for welding, grinding and paint shop. However, the potential of industrial robot applications in machining has be realized. The volume of monolithic components is large and there are many issues in machining process such as geometric tolerance and quality of machined surface. In such cases the traditional CNC machine is replaced by industrial robots, which will reduce the production cost, reduce labor and increase the efficiency. In this paper, the milling experiment of 7050-T7451 aeronautical aluminum alloy was carried out by using industrial robot KR210 R2700. In addition, the experiment was employed to study the influence of milling speed, feed-rate, cutting depth and cutting width on vibrations, surface roughness was also measured to evaluate the machining quality. Besides, the axis of angle was changed which led to the different industrial robot’s postures. The vibration signal of different postures was acquired, which was used to analysis the optimal workspace of industrial robot. The best process parameters were obtained, which will play a guiding significance on the actual production.

Fractal Pattern Recognition of Image Profiles for Manufacturing Process Monitoring and Control

Nowadays manufacturing industry faces increasing demands to customize products according to personal needs. This trend leads to a proliferation of complex product designs. To cope with this complexity, manufacturing systems are equipped with advanced sensing capabilities. However, traditional statistical process control methods are not concerned with the stream of in-process imaging data. Also, very little has been done to investigate nonlinearity, irregularity, and inhomogeneity in image stream collected from manufacturing processes. This paper presents the multifractal spectrum and lacunarity measures to characterize irregular and inhomogeneous patterns of image profiles, as well as detect the hidden dynamics of the underlying manufacturing process. Experimental studies show that the proposed method not only effectively characterizes the surface finishes for quality control of ultra-precision machining but also provides an effective model to link process parameters with fractal characteristics of in-process images acquired from additive manufacturing. This, in turn, will allow a swift response to processes changes and consequently reduce the number of defective products. The proposed fractal method has strong potentials to be applied for process monitoring and control in a variety of domains such as ultra-precision machining, additive manufacturing, and biomanufacturing.