Dr. Jiaming Fu

The versatility of robotic grippers enables them to handle various objects in diverse applications, primarily when operations need flexible handling of different materials and shapes. The research introduces a new grasping apparatus named Compliant Self-Adaptive Variable Stiffness Robotic Gripper (CS-VSRG) that unifies shape-conforming and adaptive compliance features to improve handling operations. Key to this innovation is the design of Self-Adaptive Variable Stiffness fingers which are constructed by the combination of three separate layers for high-stiffness outer and medium-stiffness middle and low-stiffness inner components. Each layer has been engineered to choose specific objects when gripping forces are applied. Fragile objects with a small grasping force are needed only to contact the low-stiffness layer. As the required grasping force increases, the deformation leads to the contact layer extending to the middle or outer layers through self-adaptive stiffness transition. This allows for adapting to various types of objects while maintaining a firm grasp. A comprehensive control strategy and system design are also developed and analyzed in this paper. Finite element analysis (FEA) simulations are also performed to validate the gripper. The stress distribution, deformation characteristics, and stiffness regulation of the layered structure are studied. A physical prototype is fabricated and an experimental grasping demonstration is performed to evaluate the actual performance. The results confirm that the gripper can dynamically adjust the stiffness of the finger layers involved in the grasping process, thereby enhancing both adaptability and robustness. Additionally, the proposed design is characterized by low cost, high reliability, and structural simplicity, making it well-suited for large-scale industrial applications that require cost-effective robotic operations. It can also be integrated as a component of a dexterous robotic hand.

Force feedback in compliant robotic grippers is essential for precise manipulation tasks but remains challenging because irregular deformation of soft materials renders traditional sensor integration impractical. Event camera has emerged as a powerful alternative to conventional vision sensors through their ability to capture only temporal changes in a scene, thereby significantly reducing data redundancy and processing overhead. In this paper, we introduce a novel vision-based force prediction framework that employs a Mamba-Like architecture to process event camera data in compliant grippers. To validate our approach, we have developed a custom gripper with variable stiffness and created a comprehensive dataset comprising over 9,000 event frames. Our methodology combines self-supervised pre-training for learning rich feature representations with a Mamba-like regression framework to achieve accurate force prediction. The proposed method demonstrates a 0.14 improvement in RMSE when compared to existing Vision Transformer approaches. Through extensive experimental validation-including real-time performance analysis, ablation studies, and generalization tests across various gripper configurations-we demonstrate the framework's effectiveness. Our results indicate robust performance suitable for practical industrial applications, suggesting potential extensions to other compliant robotics applications that require precise force estimation.

The construction industry heavily uses many different categories of heavy equipment on a daily basis. Generally, such equipment is used extensively and lasts a long time with limited to no possibility of updating or enhancing. Especially, with the current advancements of technology, there is a need to modernize such mechanical equipment. With regard to equipment control, most equipment requires users to stay close to or directly control the equipment. In this paper, the authors propose an upgrading system that allows the remote operation of heavy equipment through push and pull mechanism. By utilizing a single-board computer such as Raspberry Pi with linear actuators, the system can allow the users to control the heavy equipment remotely. The proposed system is successfully demonstrated by upgrading an indoor tower crane in the D. Dorsey Moss Construction Lab at Purdue University. This type of upgrade could support other heavy equipment in the construction field through integrating various sensors and actuators for digital enhancement.

The demand for flexible grasping of various objects by robotic hands in the industry is rapidly growing. To address this, we propose a novel variable stiffness gripper (VSG). The VSG design is based on a parallel-guided beam structure inserted by a slider from one end, allowing stiffness variation by changing the length of the parallel beams participating in the system. This design enables continuous adjustment between high compliance and high stiffness of the gripper fingers, providing robustness through its mechanical structure. The linear analytical model of the deflection and stiffness of the parallel beam is derived, which is suitable for small and medium deflections. The contribution of each parameter of the parallel beam to the stiffness is analyzed and discussed. Also, a prototype of the VSG is developed, achieving a stiffness ratio of 70.9, which is highly competitive. Moreover, a vision-based force sensing method utilizing ArUco markers is proposed as a replacement for traditional force sensors. By this method, the VSG is capable of closed-loop control during the grasping process, ensuring efficiency and safety under a well-defined grasping strategy framework. Experimental tests are conducted to emphasize the importance and safety of stiffness variation. In addition, it shows the high performance of the VSG in adaptive grasping for asymmetric scenarios and its ability to flexible grasping for objects with various hardness and fragility. These findings provide new insights for future developments in the field of variable stiffness grippers.

Stroke patients have a pressing demand for innovative hand rehabilitation devices with multiple finger joint positions. In this research, a novel soft variable stiffness actuator is proposed for alternative hand rehabilitation therapies. The design allows targeted stiffness modifications above specific joints, enabling a range of motions from a simple actuation source, supplemented by a smaller actuator determining internal stiffness. Four configurations were tested, adjusting stiffness regions within the actuator. Initial testing employed Finite Element Analysis simulations with hyper-elastic non-linear material settings, predicting actuator behavior under varying pressure. Comparison of simulation and experimental results, despite differing actuation pressure due to air leaks, revealed an acceptable replication of Finite Element Analysis simulations, with minor differences in tip trajectory. Validation involved testing a single prototype, laying the groundwork for a new form of variable stiffness actuator. The results indicate that the proposed soft variable stiffness actuator can achieve diverse hand rehabilitation guidance postures under the desired force, making it applicable to next-generation hand rehabilitation gloves and other devices.

Robotic grippers are receiving increasing attention in various industries as essential components of robots for interacting and manipulating objects. While significant progress has been made in the past, conventional rigid grippers still have limitations in handling irregular objects and can damage fragile objects. We have shown that soft grippers offer deformability to adapt to a variety of object shapes and maximize object protection. At the same time, dynamic vision sensors (e.g., event-based cameras) are capable of capturing small changes in brightness and streaming them asynchronously as events, unlike RGB cameras, which do not perform well in low-light and fast-moving environments. In this paper, a dynamic-vision-based algorithm is proposed to measure the force applied to the gripper. In particular, we first set up a DVXplorer Lite series event camera to capture twenty-five sets of event data. Second, motivated by the impressive performance of the Vision Transformer (ViT) algorithm in dense image prediction tasks, we propose a new approach that demonstrates the potential for force estimation and meets the requirements of real-world scenarios. We extensively evaluate the proposed algorithm on a wide range of scenarios and settings, and show that it consistently outperforms recent approaches.

Variable Stiffness Grippers (VSGs) represent a groundbreaking advancement in robotic manipulation, embodying the seamless integration of flexibility and rigidity to meet the multifaceted challenges of modern automation. These devices leverage the adaptability of compliant modes for handling a wide range of objects yet can switch to a rigid mode for tasks requiring high strength and precision. The management of variable stiffness poses significant challenges, especially in achieving precise control over the gripper's adaptability to objects of varying compliance. This paper proposes a method to provide a combination of position control and force control of a VSG by exploiting the dynamic model and the different stiffness levels. Our research examines active disturbance rejection control (ADRC) and deterministic robust control (DRC), demonstrating their advantages over PID in managing stiffness variations in robotic grippers. We highlight ADRC and DRC's enhanced robustness and adaptability through a comparative analysis, at different stiffness levels and grasping processes. These efforts highlight the importance of sophisticated control systems, in distinguishing between stiff and rigid modes effectively, enabling VSGs to handle objects ranging from fragile.

In Virtual Reality (VR), adversarial attack remains a significant se-curity threat. Most deep learning-based methods for physical and digital adversarial attacks focus on enhancing attack performance by crafting adversarial examples that contain large printable distortions that are easy for human observers to identify. However, attackers rarely impose limitations on the naturalness and comfort of the appearance of the generated attack image, resulting in a no-ticeable and unnatural attack. To address this challenge, we propose a framework to incorporate style transfer to craft adversarial inputs of natural styles that exhibit minimal detectability and maximum natural appearance, while maintaining superior attack capabilities.

Variable stiffness manipulators balance the trade-off between manipulation performance needing high stiffness and safe human-robot interaction desiring low stiffness. Variable stiffness links enable this flexible manipulation function during human-robot interaction. In this paper, we propose a novel variable stiffness link based on discrete variable stiffness units (DSUs). A DSU is a parallel guided beam that can adjust stiffness discretely by changing the cross-sectional area properties of the hollow beam segments. The variable stiffness link (Tri-DSU) consists of three tandem DSUs to achieve eight stiffness modes and a stiffness ratio of 31. To optimize the design, stiffness analysis of the DSU and Tri-DSU under various configurations and forces was performed by a derived linear analytical model which applies to small/intermediate deflections. The model is derived using the approach of serially connected beams and superposition combinations. 3D-Printed prototypes were built to verify the feature and performance of the Tri-DSU in comparison with the finite element analysis and analytical model results. It's demonstrated that our model can accurately predict the stiffnesses of the DSU and Tri-DSU within a certain range of parameters. Impact tests were also conducted to validate the performance of the Tri-DSU. The developed method and analytical model are extendable to multiple DSUs with parameter configurations to achieve modularization and customization, and also provide a tool for the design of reconfigurable collaborative robot (cobot) manipulators.

Large deflection modeling is a crucial field of study in the analysis and design of compliant mechanisms (CM). This paper proposes a machine learning (ML) approach for predicting the deflection of discrete variable stiffness units (DSUs) that cover a range from small to large deflections. The primary structure of a DSU consists of a parallel guide beam with a hollow cavity that can change stiffness discretely by inserting or extracting a solid block. The principle is based on changing the cross-sectional area properties of the hollow section. Prior to model training, a large volume of data was collected using finite element analysis (FEA) under different loads and various dimensional parameters. Additionally, we present three widely used machine learning-based models for predicting beam deflection, taking into account prediction accuracy and speed. Several experiments are conducted to evaluate the performance of the ML models that were compared with the FEA and analytical model results. The optimal ML model, multilayer perceptron (MLP), can achieve a 7.9% maximum error compared to FEA. Furthermore, the model was employed in a practical application for inverse design, with various cases presented depending on the number of solved variables. This method provides a innovative perspective for studying the modeling of compliant mechanisms and may be extended to other mechanical mechanisms.