Search for: All records

This paper presents polycarbonate negative topographies used as substrates for the templated self- assembly of microsphere-based microrobots. This approach protects primary structures from damage during molding and de-molding, providing high fidelity negatives of arrays for assembly via templated assembly by selective removal (TASR). We show that reducing the surface energy mismatch between the microspheres and substrate results in yield increases up to 790%. This work addresses yield-related challenges of multicomponent microsystem assembly with existing PDMS-based templated assembly methods. The application of this technology in DNA microswimmer fabrication is demonstrated. 

Magnetically-actuated swimming microrobots are an emerging tool for navigating and manipulating materials in confined spaces. Recent work has demonstrated that it is possible to build such systems at the micro and nanoscales using polymer microspheres, magnetic particles and DNA nanotechnology. However, while these materials enable an unprecedented ability to build at small scales, such systems often demonstrate significant polydispersity resulting from both the material variations and the assembly process itself. This variability makes it difficult to predict, let alone optimize, the direction or magnitude of microswimmer velocity from design parameters such as link shape or aspect ratio. To isolate questions of a swimmer's design from variations in its physical dimensions, we present a novel experimental platform using two-photon polymerization to build a two-link, buoyant milliswimmer with a fully customizable shape and integrated flexible linker (the swimmer is underactuated, enabling asymmetric cyclic motion and net translation). Our approach enables us to control both swimming direction and repeatability of swimmer performance. These studies provide ground truth data revealing that neither the first order nor second order models currently capture the key features of milliswimmer performance. We therefore use our experimental platform to develop design guidelines for tuning the swimming speeds, and we identify the following three approaches for increasing speed: (1) tuning the actuation frequency for a fixed aspect ratio, (2) adjusting the aspect ratio given a desired range of operating frequencies, and (3) using the weaker value of linker stiffness from among the values that we tested, while still maintaining a robust connection between the links. We also find experimentally that spherical two-link swimmers with dissimilar link diameters achieve net velocities comparable to swimmers with cylindrical links, but that two-link spherical swimmers of equal diameter do not. 

Methods for state estimation that rely on visual information are challenging on legged robots due to rapid changes in the viewing angle of onboard cameras. In this work, we show that by leveraging structure in the way that the robot locomotes, the accuracy of visual-inertial SLAM in these challenging scenarios can be increased. We present a method that takes advantage of the underlying periodic predictability often present in the motion of legged robots to improve the performance of the feature tracking module within a visual-inertial SLAM system. Our method performs multi-session SLAM on a single robot, where each session is responsible for mapping during a distinct portion of the robot’s gait cycle. Our method produces lower absolute trajectory error than several state-of-the-art methods for visual-inertial SLAM in both a simulated environment and on data collected on a quadrupedal robot executing dynamic gaits. On real-world bounding gaits, our median trajectory error was less than 35% of the error of the next best estimate provided by state-of-the-art methods. 

The complexity associated with the control of highly-articulated legged robots scales quickly as the number of joints increases. Traditional approaches to the control of these robots are often impractical for many real-time applications. This work thus presents a novel sampling-based planning ap- proach for highly-articulated robots that utilizes a probabilistic graphical model (PGM) to infer in real-time how to optimally modify goal-driven, locomotive behaviors for use in closed-loop control. Locomotive behaviors are quantified in terms of the parameters associated with a network of neural oscillators, or rather a central pattern generator (CPG). For the first time, we show that the PGM can be used to optimally modulate different behaviors in real-time (i.e., to select of optimal choice of parameter values across the CPG model) in response to changes both in the local environment and in the desired control signal. The PGM is trained offline using a library of optimal behaviors that are generated using a gradient-free optimization framework. 

Sampling-based motion planning algorithms provide a means to adapt the behaviors of autonomous robots to changing or unknown a priori environmental conditions. However, as the size of the space over which a sampling-based approach needs to search is increased (perhaps due to considering robots with many degree of freedom) the computational limits necessary for real-time operation are quickly exceeded. To address this issue, this paper presents a novel sampling-based approach to locomotion planning for highly-articulated robots wherein the parameters associated with a class of locomotive behaviors (e.g., inter-leg coordination, stride length, etc.) are inferred in real-time using a sample-efficient algorithm. More specifically, this work presents a data-based approach wherein offline-learned optimal behaviors, represented using central pattern generators (CPGs), are used to train a class of probabilistic graphical models (PGMs). The trained PGMs are then used to inform a sampling distribution of inferred walking gaits for legged hexapod robots. Simulated as well as hardware results are presented to demonstrate the successful application of the online inference algorithm. 

Sampling-based motion planning algorithms provide a means to adapt the behaviors of autonomous robots to changing or unknown a priori environmental conditions. However, as the size of the space over which a sampling-based approach needs to search is increased (perhaps due to considering robots with many degree of freedom) the computational limits necessary for real-time operation are quickly exceeded. To address this issue, this paper presents a novel sampling-based approach to locomotion planning for highly-articulated robots wherein the parameters associated with a class of locomotive behaviors (e.g., inter-leg coordination, stride length, etc.) are inferred in real-time using a sample-efficient algorithm. More specifically, this work presents a data-based approach wherein offline-learned optimal behaviors, represented using central pattern generators (CPGs), are used to train a class of probabilistic graphical models (PGMs). The trained PGMs are then used to inform a sampling distribution of inferred walking gaits for legged hexapod robots. Simulated as well as hardware results are presented to demonstrate the successful application of the online inference algorithm. 

Snake robots have the potential to locomotethrough tightly packed spaces, but turning effectively withinunmodelled and unsensed environments remains challenging.Inspired by a behavior observed in the tiny nematode wormC.elegans, we propose a novel in-place turning gait for elongatedlimbless robots. To simplify the control of the robots’ many in-ternal degrees-of-freedom, we introduce a biologically-inspiredtemplate in which two co-planar traveling waves are superposedto produce an in-plane turning motion, theomega turn. Theomega turn gait arises from modulating the wavelengths andamplitudes of the two traveling waves. We experimentally testthe omega turn on a snake robot, and show that this turninggait outperforms previous turning gaits: it results in a largerangular displacement and a smaller area swept by the bodyover a gait cycle, allowing the robot to turn in highly confinedspaces.