Michael Yu Wang

• Professor
• Professor (Full) at National University of Singapore

This survey provides a comprehensive review on recent advancements of generative learning models in robotic manipulation, addressing key challenges in the field. Robotic manipulation faces critical bottlenecks, including significant challenges in insufficient data and inefficient data acquisition, long-horizon and complex task planning, and the mul...

Ensuring safe driving while maintaining travel efficiency for autonomous vehicles in dynamic and occluded environments is a critical challenge. This paper proposes an occlusion-aware contingency safety-critical planning approach for real-time autonomous driving in such environments. Leveraging reachability analysis for risk assessment, forward reac...

In this work, we present FRTree planner, a novel robot navigation framework that leverages a tree structure of free regions, specifically designed for navigation in cluttered and unknown environments with narrow passages. The framework continuously incorporates real-time perceptive information to identify distinct navigation options and dynamically...

In this work, we propose a trajectory optimization approach for robot navigation in cluttered 3D environments. We represent the robot's geometry as a semialgebraic set defined by polynomial inequalities such that robots with general shapes can be suitably characterized. We exploit the collision-free space directly to construct a graph of free regio...

In this work, we present FRTree planner, a novel robot navigation framework that leverages a tree structure of free regions, specifically designed for navigation in cluttered and unknown environments with narrow passages. The framework continuously incorporates real-time perceptive information to identify distinct navigation options and dynamically...

Lattice materials are an emerging family of advanced engineering materials with unique advantages for lightweight applications. However, the mechanical behaviors of lattice materials at ultra‐low relative densities are still not well understood, and this severely limits their lightweighting potential. Here, a high‐precision micro‐laser powder bed f...

Large-scale 3D mapping is an important task for robotics and autonomous driving. However, mobile robots and autonomous vehicles with limited hardware resources may face issues with large memory consumption. It is challenging to achieve a balance between mapping quality and memory consumption. To address this issue, we propose a new compact implicit...

Hierarchical triply periodic minimal surface (TPMS) shell lattices exhibit a wide range of tailorable properties that can be fine-tuned by lattice design. Recent advances in additive manufacturing towards high resolution enable the precise fabrication of hierarchical TPMS lattices. However, the isotropic behavior of hierarchical TPMS lattices has r...

In real-world object manipulation scenarios, multiple mobile manipulator systems may suffer from disturbances and asynchrony, leading to excessive interaction forces and causing object damage or emergency stops. This paper presents a novel distributed motion control approach aimed at reducing these unnecessary interaction forces. The control strate...

Elastically isotropic open-cell lattice metamaterials exhibit identical elastic properties along arbitrary directions, and are ideal candidates for applications with unknown primary loading directions. Their open-cell properties are preferred for additive manufacturing processes and multifunctional applications requiring mass and heat transfer. Thi...

Enforcing safety while preventing overly conservative behaviors is essential for autonomous vehicles to achieve high task performance. In this paper, we propose a barrier-enhanced parallel homotopic trajectory optimization (BPHTO) approach with the over-relaxed alternating direction method of multipliers (ADMM) for real-time integrated decision-mak...

3D point cloud maps are widely used in robotic tasks like localization and planning. However, dynamic objects, such as cars and pedestrians, can introduce ghost artifacts during the map generation process, leading to reduced map quality and hindering normal robot navigation. Online dynamic object removal methods are restricted to utilize only local...

In-hand manipulation is challenging in robotics due to the intricate contact dynamics and high degrees of control freedom. Precise manipulation with high accuracy often requires tactile perception, which adds further complexity to the system. This paper addresses these challenges in perception and control and proposes a framework for learning high-...

This research was focused on innovative design of lattice metamaterials that can exhibit tunable double-negative mechanical properties and elastic isotropy simultaneously. A discrete topology optimization method using a multi-material ground structure was developed to create microlattices exhibiting both negative thermal expansion coefficient and n...

This letter introduces ERRA, an embodied learning architecture that enables robots to jointly obtain three fundamental capabilities (reasoning, planning, and interaction) for solving long-horizon language-conditioned manipulation tasks. ERRA is based on tightly-coupled probabilistic inferences at two granularity levels. Coarse-resolution inference...

Incorporating the process-induced material anisotropy into design framework of additive manufactured lattice structures is crucial to ensure the accuracy of design and analysis models. This work proposes an analytical approach to design stretching-dominated truss lattices with tailored elastic properties, including isotropic elasticity, tailored ze...

This letter introduces ERRA, an embodied learning architecture that enables robots to jointly obtain three fundamental capabilities (reasoning, planning, and interaction) for solving long-horizon language-conditioned manipulation tasks. ERRA is based on tightly-coupled probabilistic inferences at two granularity levels. Coarse-resolution inference...

Humans excel in grasping objects through diverse and robust policies, many of which are so probabilistically rare that exploration-based learning methods hardly observe and learn. Inspired by the human learning process, we propose a method to extract and exploit latent intents from demonstrations, and then learn diverse and robust grasping policies...

This paper tackles the task of singulating and grasping paper-like deformable objects. We refer to such tasks as paper-flipping. In contrast to manipulating deformable objects that lack compression strength (such as shirts and ropes), minor variations in the physical properties of the paper-like deformable objects significantly impact the results,...

Buckling failure is a major concern in lightweight lattice core sandwich structures (LCSSs). This study proposes an efficient method for designing LCSSs with superior buckling strength under uniaxial compression. In this approach, the buckling-resistant non-uniform shapes of single struts with different slenderness and inclination are first identif...

Multi-mobile robot systems show great advantages over one single robot in many applications. However, the robots are required to form desired task-specified formations, making feasible motions decrease significantly. Thus, it is challenging to determine whether the robots can pass through an obstructed environment under formation constraints, espec...

Vision-based tactile sensors have been widely studied in the robotics field for high spatial resolution and compatibility with machine learning algorithms. However, the currently employed sensor’s imaging system is bulky limiting its further application. Here we present a micro lens array (MLA) based vison system to achieve a low thickness format o...

Multi-mobile robot systems show great advantages over one single robot in many applications. However, the robots are required to form desired task-specified formations, making feasible motions decrease significantly. Thus, it is challenging to determine whether the robots can pass through an obstructed environment under formation constraints, espec...

Tactile sensing is an essential perception for robots to complete dexterous tasks. As a promising tactile sensing technique, vision-based tactile sensors have been developed to improve robot performance in manipulation and grasping. Here we propose a new design of a vision-based tactile sensor, DelTact. The sensor uses a modular hardware architectu...

Additively manufactured elastically isotropic lattice structures are promising for lightweight structural applications. Current design methods for elastically isotropic lattice structures are mostly based on the assumption of isotropic constitutive materials. However, additive manufacturing (AM) fabricated materials generally exhibit undesired anis...

With the growing interest in vision-based tactile sensors, various types of sensors that utilize digital imaging are being developed. Among them, a group of sensors captures the tactile flow resulted from the contact deformation using the optical flow algorithm from computer vision and achieves full resolution deformation tracking on the tactile su...

Auxetic metamaterials are a type of mechanical metamaterials possessing negative Poisson’s ratios and displaying counter-intuitive deformation behavior, which are useful for a range of potential applications. However, two limitations restrain their further development and practical application. First, most current designs are not isotropic, althoug...

Micro laser powder bed fusion (μLPBF), for the first time, enables fabrication of low-density triply periodic minimal surface (TPMS) shell lattices with smaller feature size. However, the understandings on the mechanical responses of lightweight TPMS structures by μLPBF are yet to be updated. Herein, stainless steel 316L TPMS shell lattices (i.e.,...

Multi-robot system has great potential than one single robot due to the flexibility. By adjusting formations of the system, complex tasks can be accomplished. However, formation control may conflict with obstacle avoidance in clutter environments. Therefore, we propose a distributed motion planner for multi-mobile robot systems to balance formation...

Robotic grippers with visuotactile sensors have access to rich tactile information for grasping tasks but encounter difficulty in partially encompassing large objects with sufficient grip force. While hierarchical gecko-inspired adhesives are a potential technique for bridging performance gaps, they require a large contact area for efficient usage....

With the continuous improvement in the precision of the laser powder bed fusion (LPBF) process, printing parts with highly complex geometries and fine details has been realized. Given the large design freedom provided by LPBF, various shell lattice structures have been designed and fabricated successfully. However, few works focus on predicting and...

Controllable dry adhesives have attracted significant interest; however, their development and application remain limited owing to their high cost, complexity, and machine precision required for their fabrication. Herein, a scalable roll‐to‐roll technique for fabricating wedge‐shaped controllable dry adhesives for use in robotic grasping tasks is d...

Robotic grippers with visuotactile sensors have access to rich tactile information for grasping tasks but encounter difficulty in partially encompassing large objects with sufficient grip force. While hierarchical gecko-inspired adhesives are a potential technique for bridging performance gaps, they require a large contact area for efficient usage....

Vision-based tactile sensors have been widely studied in the robotics field for high spatial resolution and compatibility with machine learning algorithms. However, the currently employed sensor's imaging system is bulky limiting its further application. Here we present a micro lens array (MLA) based vison system to achieve a low thickness format o...

Tactile sensing is an essential perception for robots to complete dexterous tasks. As a promising tactile sensing technique, vision-based tactile sensors have been developed to improve robot manipulation performance. Here we propose a new design of our vision-based tactile sensor, DelTact, with its high-resolution sensing abilities of multiple moda...

We present a family of elastically isotropic stretching-dominated cubic truss lattices. Their Young’s, shear, bulk moduli approach nearly 1/3 of the theoretical Hashin-Shtrikman upper bounds at low density limit, which are the maximum isotropic stiffness that truss lattices can achieve. The proposed lattices are designed by combination of two eleme...

Designs of soft actuators are mostly guided and limited to certain target functionalities. This article presents a novel programmable design for soft pneumatic bellows-shaped actuators with distinct motions, thus a wide range of functionalities can be engendered through tuning channel parameters. According to the design principle, a kinematic model...

In this paper, an adaptive parameterized level set topology optimization method (APLSM) is proposed by using a bilinear basis function which is common in the finite element method (FEM) to parameterize the level set function. A CPU parallel computing strategy is applied for large-scale topology optimization problems. In the presented APLSM, the sol...

Humans excel in grasping objects through diverse and robust policies, many of which are so probabilistically rare that exploration-based learning methods hardly observe and learn. Inspired by the human learning process, we propose a method to extract and exploit latent intents from demonstrations, and then learn diverse and robust grasping policies...

Artificial tactile sensing for robots is a counterpart to the human sense of touch, serving as a feedback interface for sensing and interacting with the environment. A vision-based tactile sensor has emerged as a novel and advantageous branch of artificial tactile sensors. Compared with conventional tactile sensors, vision-based tactile sensors pos...

Triply periodic minimal surface (TPMS) shell lattices are attracting increasingly attention due to their unique combination of geometric and mechanical properties, and their open-cell topology. However, uniform thickness TPMS shell lattices are usually anisotropic in stiffness, namely having different Young’s moduli along different lattice directio...

Lattices with hierarchical architectures exhibit unique geometric and mechanical properties compared with single scale ones. While numerous research efforts have focused on hierarchical strut lattices, hierarchical sheet lattices have yet to be studied in detail. This paper proposes a systemic framework including geometric design, finite element mo...

Triply periodic minimal surface (TPMS) shell-lattices are attracting increasingly attention due to their unique combination of geometric and mechanical properties, and their open-cell topology. However, uniform thickness TPMS shell-lattices are usually anisotropic in stiffness, namely having different Young's moduli along different lattice directio...

A parametric level set approach based on B-splines is developed for the stress-based shape and topology optimization of cellular structures. In this method, the whole design domain is divided into a set of non-overlapping sub-domains, within each of which the structure is represented by an implicit B-spline level set function. This parameterization...

Robotics has undergone a profound revolution in the past 50 years, moving from the laboratory and research institute to the factory and home. Kinematics and dynamics theories have been developed as the foundation for robot design and control, based on the conventional definition of robots: a kinematic chain of rigid links.

The level set method (LSM), which is transplanted from the computer graphics field, has been successfully introduced into the structural topology optimization field for about two decades, but it still has not been widely applied to practical engineering problems as density-based methods do. One of the reasons is that it acts as a boundary evolution...

In this work, we propose a new kind of soft-rigid hybrid actuator composed mainly of soft chambers and rigid frames. Compared with the well-known fiber-reinforced soft actuators, the hybrid actuators are able to ensure the design of noncircular cross-sectional shapes. It is demonstrated that rigid frames are capable of providing geometric constrain...

A novel subdomain structural topology optimization method is proposed for the minimum compliance problem based on the level sets with the parameterization of radial basis function (RBF). In this method, the level set function evolves on each subdomain separately and independently according to the requirements of objective functions and additional c...

Additive manufacturing enables the fabrication of periodic ceramic lattices with controllable micro-architectures. Many studies reported their catastrophic brittle fracture behaviour. However, ceramic lattices may fail by a layer-by-layer pseudo-ductile fracture mode, by controlling micro-architectures and porosities. Moreover, their fracture behav...

Inchworms have been one of the most widely used bionic templates for designing soft robotic devices. Bioresearch has shown that muscles of inchworms exhibit nonlinear hysteresis and their body structures are with hydrostatic skeleton. But effects of these properties on their dynamic movements have not been studied yet. In this work, a dynamic model...

The level set method, which is transplanted from the computer vision field, has been successfully introduced into the structural topology optimization field for about two decades, but it is still not well implemented in the industry. In this paper, a level set band method is proposed to improve the continuity of the objective and constraint functio...

This paper presents a variable cutting (VCUT) level set based topology optimization method to design functionally graded cellular structures (FGCS). A variable and continuous cutting function by interpolating with a set of height variables is proposed to generate functionally graded cellular structures, which offers a novel tool to optimize the mac...

Brain-computer interface (BCI) technology shows potential for application to motor rehabilitation therapies that use neural plasticity to restore motor function and improve quality of life of stroke survivors. However, it is often difficult for BCI systems to provide the variety of control commands necessary for multi-task real-time control of soft...

In this work, we develop a level set modeling technique for designing and optimization of solid/cellular structures, called cellular level set in B-splines (CLIBS). In this technique, the entire design domain for the solid/cellular structure in question is subdivided into a set of connected volumetric cells in three dimensions. In the level set sch...

This paper outlines a precision motion stage actuated by using a voice coil motor with a floating stator. For getting good performance, a multi-mode sliding mode control (MMSMC) was designed to operate this linear motion stage. MMSMC contains two sliding mode controllers: a sliding-mode control (SMC) and an integral sliding-mode control. The switch...

Multimaterial soft robots, composing of integrated soft actuators and a relatively harder body, show great potential to exert higher payloads and support their own weights. This paper proposed a systemic framework to automatically design and fabricate this kind of robots. The multimaterial design problem is mathematically modeled under the framewor...

This paper concerns hysteresis and creep of a piezo-actuated nano positioning stage. The hysteresis and creep, considered as bounded disturbance or uncertainty, are suppressed without a nonlinear model. An improved Fuzzy Adaptive Sliding Mode Control (FASMC) with a Proportional-Integral-Derivative sliding surface is designed to cancel both hysteres...

Designing metallic cellular structures with triply periodic minimal surface (TPMS) sheet cores is a novel approach for lightweight, multi-functional structural applications. Different from current honeycombs and lattices, TPMS sheet structures are composed of continuous and smooth shells, allowing for large surface areas and continuous internal cha...

This paper presents a compact and efficient 88-line MATLAB code for the parameterized level set method based topology optimization using radial basis functions (RBFs), which is applied to minimize the compliance of a two-dimensional linear elastic structure. This parameterized level set method using radial basis functions can maintain a relatively...

Soft robots, are mobile machines largely constructed from soft materials and have received much attention recently because they are opening new perspectives for robot design and control. This paper reports recent progress in the development of soft robots, more precisely, soft actuators and soft sensors. Soft actuators play an important role in fun...

In tapping-mode atomic force microscopy, the higher harmonics generated in the tapping process provide evidence for material composition imaging based on material property information (e.g. elasticity). But problems of low amplitude and rapid decay of higher harmonics restrict their sensitivity and accessibility. The probe’s characteristic of assig...

This paper presents the control performance of a linear motion stage driven by Voice Coil Motor (VCM). Unlike the conventional VCM, the stator of this VCM is regulated, which means it can be adjusted as a floating-stator or fixed-stator. A Multi-Mode Sliding Mode Control (MMSMC), including a conventional Sliding Mode Control (SMC) and an Integral S...

In multifrequency atomic force microscopy(AFM), probe’s characteristic of assigning resonance frequencies to integer harmonics results in a remarkable improvement of detection sensitivity at specific harmonic components. The selection criterion of harmonic order is based on its amplitude’s sensitivity on material properties, e.g., elasticity. Previ...

Macro–micro motion stage is mainly applied in microelectronics manufacturing to realize a high-acceleration, high-speed and nano-positioning motion. The high acceleration and nano-positioning accuracy would be influenced by the vibration of the motion stage. In the paper, a concept of floating stage is introduced in the macro–micro motion for isola...

To take into account the shrinkage effect in the early stage of Reinforced Concrete (RC) design, an effective continuum topology optimization method is presented in this paper. Based on the power-law interpolation, shrinkage of concrete is numerically simulated by introducing an additional design-dependent force. Under multi-axial stress conditions...

In this paper, flexural vibration in a locally-resonant (LR) beam with periodically attached separated force and moment beam-like resonators is investigated theoretically and experimentally. The relationship between the distance parameter and the band structure of an Euler-Bernoulli beam with proposed locally resonators is provided using the transf...

Connection frame is an essential component to implement high acceleration and ultra-precision positioning motion in a macro–micro motion platform. The performance of the positioning system is mainly affected by two sources which include thermal–mechanical deformation and the natural frequency of connection frame. In the paper, multi-objective optim...

This paper presents a simple but effective additive hyperelasticity technique to circumvent numerical difficulties in solving the material density-based topology optimization of elastic structures undergoing large displacements. By adding a special hyperelastic material to the design domain, excessive distortion and numerical instability occurred i...

In multifrequency atomic force microscopy (AFM) to simultaneously measure topography and material properties of specimens, it is highly desirable that the higher order resonance frequencies of the cantilever probe are assigned to be integer harmonics of the excitation frequency. The harmonic resonances are essential for significant enhancement of t...

The topology optimization problem with pressure load is solved by using a level set method. The free boundary and the pressure boundary of a structure are represented separately as two zero-level sets of two level set functions, and they are independently propagated during the optimization by solving two Hamilton–Jacobi equations. In order to preve...

We present a level set method for the shape and topology optimization of both structure and support. Two level set functions are used to implicitly represent a structure. The traction free boundary and the Dirichlet boundary are represented separately and are allowed to be continuously changed during the optimization. The optimization problem of mi...

A novel optimization method, stiffness spreading method (SSM), is proposed for layout optimization of truss structures. In this method, stiffness matrices of the bar elements in a truss structure are represented by a set of equivalent stiffness matrices which are embedded in a weak background mesh. When the proposed method is used, it is unnecessar...

Both structural sizes and dimensional tolerances strongly influence the manufacturing cost and the functional performance of a practical product. This paper presents an optimization method to simultaneously find the optimal combination of structural sizes and dimensional tolerances. Based on a probability-interval mixed reliability model, the impre...

In an operation mode of atomic force microscopy that uses a higher eigenmode to determine the physical properties of material surface, the ratio between the eigenfrequency of a higher flexural eigenmode and that of the first flexural eigenmode was identified as an important parameter that affects the sensitivity and accessibility. Structure feature...

Engineering features are regular and simple shape units containing specific engineering significance. It is useful to combine feature design with structural optimization. This paper presents a generic method to design engineering features for level set based structural optimization. A Constructive Solid Geometry based Level Sets (CSGLS) description...

Based on the analysis of a typical configuration of piezoresistive sensors, the intensity and uniformity of stresses produced by measurand force in the piezoresistive elements are identified as the parameters that affect the sensitivity of a sensor, and stresses produced by perturbation forces in the piezoresistive elements are identified as the pa...

Following the extended two-material density penalization scheme, a stress-based topology optimization method for the layout design of prestressed concrete structures is proposed. The Drucker–Prager yield criterion is used to predict the asymmetrical strength failure of concrete. The prestress is considered by making a reasonable assumption on the p...

Higher harmonics signal in tapping mode atomic force microscope (AFM) topography imaging process is verified to be an evidence for material mapping in sample surface. Due to low amplitude and rapid decay of higher harmonics, structure design focused on cantilever probe is proposed to optimize the frequency response performance of probe. Here, we ut...

A move limit strategy is proposed for level set based structural optimization, which offers a tool to restrict the allowable change of the free boundary of a structure in each iterative step of optimization. To realize the move limit strategy, a function for modifying the extended design velocity is proposed. Application of the move limit strategy...

Simulation based structural design usually involves frequent conversions between implicit and explicit models. In this paper, a fast and robust mesh simplification method is proposed to improve the efficiency of implicit evaluation of large size engineered models. The mesh is simplified via edge collapse with a metric consisting of edge length and...