Mountain Scholar :: Browsing by Author "Zhao, Jianguo, committee member"

Browsing by Author "Zhao, Jianguo, committee member"

Now showing 1 - 9 of 9

Open Access
Characterizing the self-motion manifolds of redundant robots of arbitrary kinematic structures
(Colorado State University. Libraries, 2022) Almarkhi, Ahmad A., author; Maciejewski, Anthony A., advisor; Chong, Edwin, committee member; Oprea, Iuliana, committee member; Zhao, Jianguo, committee member
Robot fault tolerance measures can be classified into two categories: 1) Local measures that are based on the singular value decomposition (SVD) of the robot Jacobian, and 2) Global measures that are suitable to quantify the fault tolerance more effectively in pick-and-place applications. One can use the size of the self-motion manifold of a robot as a global fault-tolerance measure. The size of the self-motion manifold at a certain end-effector location can be simply the sum of the range of the joint angles of a robot at that location. This work employs the fact that the largest self-motion manifolds occur due to merging two (or more) previously disjoint manifolds. The connection of previously disjoint manifolds occur in special configurations in the joint space called singularities. Singularities (singular configurations) occur when two or more of the robot joint axes become aligned and are linearly dependent. A significant amount of research has been performed on identifying the robot singularities but was all based on symbolically solving for when the robot Jacobian is not of full rank. In this work, an algorithm was proposed that is based on the gradient of the singular values of the robot Jacobian. This algorithm is not limited to any Degree of Freedom (DoF) nor any specific robot kinematic structure and any rank of singularity. Based on the robot singularities, one can search for the largest self-motion manifold near robot singularities. The measure of the size of the self-motion manifold was chosen to eliminate the effect of the self-motion manifold's topology and dimension. Because the SVD at singularities is indistinct, one can employ Givens rotations to define the physically meaningful singular directions, i.e., the directions where the robot is not able to move. This approach has been extensively implemented on a 4-DoF robot, different 7-DoF robot, and an 8-DoF robot. The global fault-tolerance measure might be further optimized by changing the kinematic structure of a robot. This may allow one to determine a globally fault-tolerant robot, i.e., a robot with 2π range for all of its joint angles at certain end-effector location, i.e., a location that is the most suitable for pick-and-place tasks.
Embargo
Computational methods for the analysis of cell migration and motility
(Colorado State University. Libraries, 2024) Havenhill, Eric Colton, author; Ghosh, Soham, advisor; Heyliger, Paul, committee member; McGilvray, Kirk, committee member; Zhao, Jianguo, committee member
Collective cell migration (CCM) is necessary for many biological processes, such as in the formation or regeneration of tissue, fibroblast movement in wound healing, and the movement of macrophages and neutrophils in the body's immune response, to name a few. CCM is commonly modeled with PDEs, however these equations usually model the population density, rather than the displacement field describing the movement of any arbitrary cell. One unknown aspect of this movement is the various methods that cells use to facilitate communication to each other. Chemical communication plays a substantial role in directed cell movement, however, other mechanical methods, such as the propagation of stresses through a shared substrate to neighboring cells and cell behavior in a crowded environment, also play an important role which is less understood. The quantification of the kinematic and dynamic characteristics in CCM would present several novel advancements in understanding the collective cell behavior. First, the dynamic mode decomposition (DMD) framework is utilized. DMD allows for the recovery of a dynamic system, in the form of an ODE or PDE, by sampling the states of a system. In the context of the cell migration, the displacements of fibroblasts during a scratch-wound assay are obtained, which result in a governing PDE through the DMD process. This PDE is used in conjunction with modern optimal control theory to develop a 2D and 3D trajectory for the migration of controllable cells to a target. On an individual level, with the hybrid use of modern static structural optimization and simple non-linear control, a cell's cytoskeleton during migration can be studied, providing for the quantification of the traction force exerted on the substrate. The results of this analysis are compared with stress and structural optimization models in ANSYS and FEBio, which uses the finite element method, so that a reasonable range of these stresses during CCM can be provided. To further study the individual mechanics of cell migration, the proposed hybrid model is extended to a fully dynamic model which predicts the cytoskeletal stress fiber formations over time that require the minimal amount of material with the use of optimal control theory. The results of this research could provide useful applications in many real-world situations, from the generating of a trajectory for microrobots during drug delivery to the study of the collective migration of organisms including cells.
Open Access
Enabling predictive energy management in vehicles
(Colorado State University. Libraries, 2018) Asher, Zachary D., author; Bradley, Thomas H., advisor; Chong, Edwin, committee member; Young, Peter, committee member; Zhao, Jianguo, committee member
Widespread automobile usage provides economic and societal benefits but combustion engine powered automobiles have significant economic, environmental, and human health costs. Recent research has shown that these costs can be reduced by increasing fuel economy through optimal energy management. A globally optimal energy management strategy requires perfect prediction of an entire drive cycle but can improve fuel economy by up to 30\%. This dissertation focuses on bridging the gap between this important research finding and implementation of predictive energy management in modern vehicles. A primary research focus is to investigate the tradeoffs between information sensing, computation power requirements for prediction, and prediction effort when implementing predictive energy management in vehicles. These tradeoffs are specifically addressed by first exploring the resulting fuel economy from different types of prediction errors, then investigating the level of prediction fidelity, scope, and real-time computation that is required to realize a fuel economy improvement, and lastly investigating a large computational effort scenario using only modern technology to make predictions. All of these studies are implemented in simulation using high fidelity and physically validated vehicle models. Results show that fuel economy improvements using predictive optimal energy management are feasible despite prediction errors, in a low computational cost scenario, and with only modern technology to make predictions. It is anticipated that these research findings can inform new control strategies to improve vehicle fuel economy and alleviate the economic, environmental, and human health costs for the modern vehicle fleet.
Open Access
Indoor positioning with deep learning for mobile IoT systems
(Colorado State University. Libraries, 2022) Wang, Liping, author; Pasricha, Sudeep, advisor; Kim, Ryan, committee member; Zhao, Jianguo, committee member
The development of human-centric services with mobile devices in the era of the Internet of Things (IoT) has opened the possibility of merging indoor positioning technologies with various mobile applications to deliver stable and responsive indoor navigation and localization functionalities that can enhance user experience within increasingly complex indoor environments. But as GPS signals cannot easily penetrate modern building structures, it is challenging to build reliable indoor positioning systems (IPS). Currently, Wi-Fi sensing based indoor localization techniques are gaining in popularity as a means to build accurate IPS, benefiting from the prevalence of 802.11 family. Wi-Fi fingerprinting based indoor localization has shown remarkable performance over geometric mapping in complex indoor environments by taking advantage of pattern matching techniques. Today, the two main information extracted from Wi-Fi signals to form fingerprints are Received Signal Strength Index (RSSI) and Channel State Information (CSI) with Orthogonal Frequency-Division Multiplexing (OFDM) modulation, where the former can provide the average localization error around or under 10 meters but has low hardware and software requirements, while the latter has a higher chance to estimate locations with ultra-low distance errors but demands more resources from chipsets, firmware/software environments, etc. This thesis makes two novel contributions towards realizing viable IPS on mobile devices using RSSI and CSI information, and deep machine learning based fingerprinting. Due to the larger quantity of data and more sophisticated signal patterns to create fingerprints in complex indoor environments, conventional machine learning algorithms that need carefully engineered features suffer from the challenges of identifying features from very high dimensional data. Hence, the abilities of approximation functions generated from conventional machine learning models to estimate locations are limited. Deep machine learning based approaches can overcome these challenges to realize scalable feature pattern matching approaches such as fingerprinting. However, deep machine learning models generally require considerable memory footprint, and this creates a significant issue on resource-constrained devices such as mobile IoT devices, wearables, smartphones, etc. Developing efficient deep learning models is a critical factor to lower energy consumption for resource intensive mobile IoT devices and accelerate inference time. To address this issue, our first contribution proposes the CHISEL framework, which is a Wi-Fi RSSI- based IPS that incorporates data augmentation and compression-aware two-dimensional convolutional neural networks (2D CAECNNs) with different pruning and quantization options. The proposed model compression techniques help reduce model deployment overheads in the IPS. Unlike RSSI, CSI takes advantages of multipath signals to potentially help indoor localization algorithms achieve a higher level of localization accuracy. The compensations for magnitude attenuation and phase shifting during wireless propagation generate different patterns that can be utilized to define the uniqueness of different locations of signal reception. However, all prior work in this domain constrains the experimental space to relatively small-sized and rectangular rooms where the complexity of building interiors and dynamic noise from human activities, etc., are seldom considered. As part of our second contribution, we propose an end-to-end deep learning based framework called CSILoc for Wi-Fi CSI-based IPS on mobile IoT devices. The framework includes CSI data collection, clustering, denoising, calibration and classification, and is the first study to verify the feasibility to use CSI for floor level indoor localization with minimal knowledge of Wi-Fi access points (APs), thus avoiding security concerns during the offline data collection process.
Open Access
Kinematic design and motion planning of fault tolerant robots with locked joint failures
(Colorado State University. Libraries, 2019) Xie, Biyun, author; Maciejewski, Anthony A., advisor; Chong, Edwin K. P., committee member; Pezeshki, Ali, committee member; Zhao, Jianguo, committee member
The problem of kinematic design and motion planning of fault tolerant robots with locked joint failure is studied in this work. In kinematic design, the problem of designing optimally fault tolerant robots for equal joint failure probabilities is first explored. A measure of local fault tolerance for equal joint failure probabilities has previously been defined based on the properties of the singular values of the Jacobian matrix. Based on this measure, one can determine a Jacobian that is optimal. Because these measures are solely based on the singular values of the Jacobian, permutation of the columns does not affect the optimality. Therefore, when one generates a kinematic robot design from this optimal Jacobian, there will be 7! robot designs with the same locally optimal fault tolerant property. This work shows how to analyze and organize the kinematic structure of these 7! designs in terms of their Denavit and Hartenberg (DH) parameters. Furthermore, global fault tolerant measures are defined in order to evaluate the different designs. It is shown that robot designs that are very similar in terms of DH parameters, e.g., robots generated from Jacobians where the columns are in reverse order, can have very different global properties. Finally, a computationally efficient approach to calculate the global pre- and post-failure dexterity measures is presented and used to identify two Pareto optimal robot designs. The workspaces for these optimal designs are also shown. Then, the problem of designing optimally fault tolerant robots for different joint failure probabilities is considered. A measure of fault tolerance for different joint failure probabilities is defined based on the properties of the singular values of the Jacobian after failures. Using this measure, methods to design optimally fault tolerant robots for an arbitrary set of joint failure probabilities and multiple cases of joint failure probabilities are introduced separately. Given an arbitrary set of joint failure probabilities, the optimal null space that optimizes the fault tolerant measure is derived, and the associated isotropic Jacobians are constructed. The kinematic parameters of the optimally fault tolerant robots are then generated from these Jacobians. One special case, i.e., how to construct the optimal Jacobian of spatial 7R robots for both positioning and orienting is further discussed. For multiple cases of joint failure probabilities, the optimal robot is designed through optimizing the sum of the fault tolerant measures for all the possible joint failure probabilities. This technique is illustrated on planar 3R robots, and it is shown that there exists a family of optimal robots. After the optimally fault tolerant robots are designed, the problem of planning the optimal trajectory with minimum probability of task failure for a set of point-to-point tasks, after experiencing locked joint failures, is studied. The proposed approach first develops a method to calculate the probability of task failure for an arbitrary trajectory, where the trajectory is divided into small segments, and the probability of task failure of each segment is calculated based on its failure scenarios. Then, a motion planning algorithm is proposed to find the optimal trajectory with minimum probability of task failure. There are two cases. The trajectory in the first case is the optimal trajectory from the start configuration to the intersection of the bounding boxes of all the task points. In the other case, all the configurations along the self-motion manifold of task point 1 need to be checked, and the optimal trajectory is the trajectory with minimum probability of task failure among them. The proposed approach is demonstrated on planar 2R redundant robots, illustrating the effectiveness of the algorithm.
Open Access
Material validation and part authentication process using hardness indentations with robotic arm implementation
(Colorado State University. Libraries, 2021) Weinmann, Katrina J., author; Simske, Steve, advisor; Chen, Thomas, committee member; Ma, Kaka, committee member; Zhao, Jianguo, committee member
In today's global economy, there are many levels of validation and authentication which must occur during manufacturing and distribution processes to ensure sufficient cyber-physical security of parts. This includes material inspection and validation during manufacturing, a method of track-and-trace for the entire supply chain, and individual forensic authentication of parts to prevent counterfeiting at any point in the manufacturing or distribution process. Traditionally, each level of validation or authentication is achieved through a separate step in the manufacturing or distribution process. In this work, a process is presented that uses hardness testing and the resulting indentations to simultaneously provide three critical functions for part validation and authentication: (i) material property validation and material property mapping achieved by administering multiple hardness tests over a given area on the part, (ii) part serialization that can be used for track-and-trace through administering hardness tests in a specific 'barcode' pattern, and (iii) the opportunity for forensic-level authentication through use of high-resolution images of the indents. Additionally, a fourth manufacturing advantage is gained in the provision of improved bonding potential for adhesive joints provided by the increase in surface area and surface roughness resulting from the addition of indents to the adherend surface. A methodology for implementing this process using a robotic arm with an end-effector-mounted portable hardness tester is presented. Implementation using a robotic arm allows a high degree of customizability of the process without changes in setup, making this process ideal for additive manufactured parts, which are often custom or low-batch and require a higher level of material validation. As a whole, this work presents a highly-customizable, single-step process that provides multi-level quality control, validation, authentication, and cyber-physical security of parts throughout the manufacturing and distribution processes
Open Access
Modeling, simulation, and control of soft robots using Koopman operator theory
(Colorado State University. Libraries, 2023) Singh, Ajai, author; Chong, Edwin K. P., advisor; Zhao, Jianguo, committee member; Pasricha, Sudeep, committee member
In nature, animals with soft body parts can control their parts to different shapes, e.g., an elephant trunk can wrap on a tree branch to pick it up. But most research on manipulators only focuses on how to control the end effector, partly because the arm of the manipulator is rigidly articulated. With recent advances in soft robotics research, controlling a soft manipulator into many different shapes will significantly improve the robot's functionality, such as medical robots morphing their shape to navigate the digestive system and then delivering drugs to the required location. However, controlling the shape of soft robots is challenging since the dynamics of soft robots are highly nonlinear and computationally intensive. In this research, we leverage a data-driven method using the Koopman operator to realize the shape control of soft robots. The dynamics of a soft manipulator are simulated using a physics-based simulator (PyElastica) to generate the input-output data. The data is used to identify an approximated linear model based on the Koopman operator. We then formulate the shape-control problem as a convex optimization problem that is computationally efficient. We demonstrated the linear model is over 12 times faster than the physics-based model in simulating the manipulator's motion. Further, we can control a soft manipulator into different shapes using model predictive control (MPC), and then in the subsequent chapters, we build a soft grid consisting of 40 such soft manipulators. We then address the issues related to the Extended Dynamic Mode Decomposition (EDMD) algorithm used for approximating the Koopman operator by developing a deep learning-based framework to learn the Koopman embeddings. On comparing the EDMD and deep learning framework it was found that the deep learning framework was far more accurate than the EDMD framework We then show that the proposed methods can be effectively used to control the shapes of soft robots by having the single soft manipulator morph into "C", "S", and "U" shapes and then extend the shape control method to the soft grid by morphing it into 3 different shapes. We envision that shape control will allow the soft robots to interact with uncertain environments or the shapes of shape-morphing robots to fulfill different tasks.
Open Access
Systems engineering assessment and experimental evaluation of quality paradigms in high-mix low-volume manufacturing environments
(Colorado State University. Libraries, 2023) Normand, Amanda, author; Bradley, Thomas, advisor; Miller, Erika, committee member; Vans, Marie, committee member; Zhao, Jianguo, committee member; Sullivan, Shane, committee member
This research aimed to evaluate the effectiveness of applying industrial paradigm application in high-mix low-volume manufacturing (HMLV) environments using a Systems Engineering approach. An analysis of existing industrial paradigms was conducted and then compared to a needs analysis for a specific HMLV manufacturer. Several experiments were selected for experimental evaluation, inspired by the paradigms, in a real-world HMLV manufacturing setting. The results of this research showed that a holistic approach to paradigm application is essential for achieving optimal performance, based on cost advantage, throughput, and flexibility, in the HMLV manufacturing environment. The findings of this research study provide insights into the importance of considering the entire manufacturing system, including both technical and human factors, when evaluating the effectiveness of industrial paradigms. Additionally, this research highlights the importance of considering the unique characteristics of HMLV manufacturing environments, such as the high degree of variability and frequent changes in product mix in designing manufacturing systems. Overall, this research demonstrates the value of a systems engineering approach in evaluating and implementing industrial paradigms in HMLV manufacturing environments. The results of this research provide a foundation for future research in this field and can be used to guide organizations in making informed decisions about production management practices in HMLV manufacturing environments.
Embargo
Towards automated manufacturing of composites via thermally assisted frontal polymerization
(Colorado State University. Libraries, 2024) Jordan, Walter Patrick, author; Yourdkhani, Mostafa, advisor; Zhao, Jianguo, committee member; Simske, Steve, committee member
Current methods for the manufacturing and repair of fiber-reinforced thermoset composites are energy-intensive, slow, and costly due to extensive processing steps and expensive equipment required to achieve complete cure. This is especially true for large, complex geometries that require autoclaves and prolonged cure times. As a result, there is a need to develop faster, cost-effective, energy-efficient processes. With the implementation of rapid curing thermoset resins, the cure cycle can be reduced from hours to minutes. This research focuses on the development, implementation, and testing of these resin systems in the established fields of mobile additive manufacturing and filament winding to demonstrate unprecedented, rapid manufacturing of composite parts. Additive manufacturing of fiber-reinforced thermoset composites is desirable due to its inherent ability to produce custom, complex parts quickly, with minimal required tooling. By printing and simultaneously curing the composite as it is deposited, freeform unsupported structures with high mechanical properties can be created. One limitation of current additive manufacturing methods is the print volume associated with traditional gantry style additive manufacturing systems. By combining the highly desirable properties of additive manufacturing using rapid, thermally curable resin systems with the mobility of a mobile additive manufacturing system, large, mechanically sound structures with virtually no limitations on print volume can be created. Moreover, rapid curing thermoset resin systems have the potential to revolutionize traditional composite manufacturing processes. Due to its wide range of applications and its ubiquitous nature, filament winding serves as a natural starting point to do so. Traditional filament winding is typically a two-step manufacturing process, where the composite part is first wound on a rotating mandrel and then cured using autoclaves or ovens. By combining these processes on the winding machine, the labor involved in manufacturing, the energy required for curing, and the overall production time are significantly reduced. In this research, a mobile additive manufacturing robot is designed, validated, and optimized for accurate locomotion and fast, dimensionally accurate printing of composite structures with high fiber alignment and degree of cure. The capabilities of this system are exhibited throughout several demonstrations that involve printing unsupported structures upside-down, the manufacturing of a bridge strong enough for the robot to pass over, and bridging the material across a 60 cm gap. Additionally, a pre-existing filament winding machine is optimized for the manufacturing of large, geometrically unconstrained composite structures. Improvements in fiber volume fraction are achieved through processing changes and a thermal profile for dry fibers is established to facilitate identification of frontal polymerization.

Browsing by Author "Zhao, Jianguo, committee member"

Results Per Page

Sort Options