IRIM Seminar Series

Series Type
Event Series
Associated Organization(s)
Associated Organization(s)

Publication Search Results

Now showing 1 - 10 of 26
  • Item
    Synthesizing & Guaranteeing Robot Behaviors
    (Georgia Institute of Technology, 2022-03-09) Kress-Gazit, Hadas ; Georgia Institute of Technology. Institute for Robotics and Intelligent Machines ; Cornell University. Sibley School of Mechanical and Aerospace Engineering
    In this talk I will describe how formal methods such as synthesis – automatically creating a system from a formal specification – can be leveraged to design robots, guarantee their behavior, and provide feedback about things that might go wrong. I will discuss the benefits and challenges of synthesis techniques and will give examples of different robotic systems including modular robots, swarms, and robots interacting with people.
  • Item
    Institute for Robotics and Intelligent Machines (IRIM) Industry Day Panel - Career Options in Robotics
    (Georgia Institute of Technology, 2021-03-17) Burnstein, Jeff ; Battles, Jon ; Joppru, Mark ; Balakirsky, Stephen ; Georgia Institute of Technology. Institute for Robotics and Intelligent Machines ; Association for Advancing Automation (A3) ; (Firm) ; ABB Robotics ; Georgia Tech Research Institute. Aerospace, Transportation, and Advanced Systems Laboratory
    The robotics industry is one of the fastest growing industries in the world. The industry is rapidly changing as robotics have moved from the manufacturing floor into almost every segment of our lives. This seminar will be a panel discussion with industry leaders to discuss the career opportunities in companies and the breath of opportunities available to the students. They will discuss internship opportunities as well as full-time positions with their respective companies.
  • Item
    Learning from the Field: Physically-based Deep Learning to Advance Robot Vision in Natural Environments
    (Georgia Institute of Technology, 2020-01-08) Skinner, Katherine ; Georgia Institute of Technology. Institute for Robotics and Intelligent Machines ; Georgia Institute of Technology. School of Aerospace Engineering
    Field robotics refers to the deployment of robots and autonomous systems in unstructured or dynamic environments across air, land, sea, and space. Robust sensing and perception can enable these systems to perform tasks such as long-term environmental monitoring, mapping of unexplored terrain, and safe operation in remote or hazardous environments. In recent years, deep learning has led to impressive advances in robotic perception. However, state-of-the-art methods still rely on gathering large datasets with hand-annotated labels for network training. For many applications across field robotics, dynamic environmental conditions or operational challenges hinder efforts to collect and manually label large training sets that are representative of all possible environmental conditions a robot might encounter. This limits the performance and generalizability of existing learning-based approaches for robot vision in field applications. In this talk, I will discuss my work to develop approaches for unsupervised learning to advance perceptual capabilities of robots in underwater environments. The underwater domain presents unique environmental conditions to robotic systems that exacerbate the challenges in perception for field robotics. To address these challenges, I leverage physics-based models and cross-disciplinary knowledge about the physical environment and the data collection process to provide constraints that relax the need for ground truth labels. This leads to a hybrid model-based, data-driven solution. I will also present work that relates this framework to challenges for autonomous vehicles in other domains.
  • Item
    Towards Robust HRI: A Stochastic Optimization Approach
    (Georgia Institute of Technology, 2021-03-31) Nikolaidis, Stefanos ; Georgia Institute of Technology. Institute for Robotics and Intelligent Machines ; University of Southern California
    The growth of scale and complexity of interactions between humans and robots highlights the need for new computational methods to automatically evaluate novel algorithms and applications. Exploring the diverse scenarios of interaction between humans and robots in simulation can improve understanding of complex HRI systems and avoid potentially costly failures in real-world settings. In this talk, I propose formulating the problem of automatic scenario generation in HRI as a quality diversity problem, where the goal is not to find a single global optimum, but a diverse range of failure scenarios that explore both environments and human actions. I show how standard quality diversity algorithms can discover interesting and diverse scenarios in the shared autonomy domain. I then propose a new quality diversity algorithm, CMA-ME, that achieves significantly better performance than the state-of-the-art in benchmark domains. Finally, I discuss applications in procedural content generation and human preference learning.
  • Item
    Conflict-Aware Risk-Averse and Safe Reinforcement Learning: A Meta-Cognitive Learning Framework
    (Georgia Institute of Technology, 2021-04-14) Modares, Hamidreza ; Georgia Institute of Technology. Institute for Robotics and Intelligent Machines ; Michigan State University. Department of Mechanical Engineering
    While the success of reinforcement learning (RL) in computer games has shown impressive engineering feat, unlike the computer games, safety-critical settings such as unmanned vehicles must thrash around in the real world, which makes the entire enterprise unpredictable. Standard RL practice generally implants pre-specified performance metrics or objectives into the RL agent to encode the designers’ intention and preferences in achieving different and sometimes conflicting goals (e.g., cost efficiency, safety, speed of response, accuracy, etc.). Optimizing pre-specified performance metrics, however, cannot provide safety and performance guarantees across a vast variety of circumstances that the system might encounter in non-stationary and hostile environments. In this talk, I will discuss novel metacognitive RL algorithms to learn not only a control policy that optimizes accumulated reward values, but also what reward functions to optimize in the first place to formally assure safety with a good enough performance. I will present safe RL algorithms that adapt the focus of attention of RL algorithm to its variety of performance and safety objectives to resolve conflict and thus assure the feasibility of the reward function in a new circumstance. Moreover, model-free RL algorithms will be presented to solve the risk-averse optimal control (RAOC) problem to optimize the expected utility of outcomes while reducing the variance of cost under aleatory uncertainties (i.e., randomness). This is because, performance-critical systems must not only optimize the expected performance, but also reduce its variance to avoid performance fluctuation during RL’s course of operation. To solve the RAOC problem, I will present the three variants of RL algorithms and analyze their advantages and preferences for different situations/systems: 1) a one-shot static convex program based RL, 2) an iterative value iteration algorithm that solves a linear programming optimization at each iteration, and 3) an iterative policy iteration algorithm that solves a convex optimization at each iteration and guarantees the stability of the consecutive control policies.
  • Item
    Understanding and Mitigating Bias in Vision Systems
    (Georgia Institute of Technology, 2021-10-06) Hoffman, Judy ; Georgia Institute of Technology. Institute for Robotics and Intelligent Machines ; Georgia Institute of Technology. School of Interactive Computing
    As visual recognition models are developed across diverse applications; we need the ability to reliably deploy our systems in a variety of environments. At the same time, visual models tend to be trained and evaluated on a static set of curated and annotated data which only represents a subset of the world. In this talk, I will then cover techniques for transferring information between different visual environments and across different semantic tasks thereby enabling recognition models to generalize to previously unseen worlds, such as from simulated to real-world driving imagery. Finally, I'll touch on the pervasiveness of dataset bias and how this bias can adversely affect underrepresented subpopulations.
  • Item
    Insect Scale Multifunctional Micro-Aerial-Robots Powered by Soft Artificial Muscles
    (Georgia Institute of Technology, 2022-09-21) Chen, Yufeng (Kevin) ; Georgia Institute of Technology. Institute for Robotics and Intelligent Machines ; Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
    Recent advances in microrobotics have demonstrated remarkable locomotive capabilities such as hovering flights, impulsive jumps, and fast running in insect-scale robots. However, most microrobots that are powered by power-dense rigid actuators have not achieved insect-like collision resilience. In this talk, I will present our recent effort in developing a new class of microrobots – ones that are powered by high bandwidth soft actuators and equipped with rigid appendages for effective interactions with environments. Towards improving collision robustness of micro-aerial robots, we develop the first heavier-than-air aerial robot powered by soft artificial muscles that demonstrates a 40-second hovering flight. In addition, our robot can recover from an in-flight collision and perform a somersault within 0.16 seconds. The robot’s maximum lift is comparable to that of the best rigid-powered sub-gram robots. This work demonstrates for the first time that soft aerial robots can achieve agile and robust flight capabilities absent in rigid-powered micro-aerial vehicles, thus showing the potential of a new class of hybrid soft-rigid robots.
  • Item
    Robot Learning: Quo Vadis?
    (Georgia Institute of Technology, 2020-11-18) Peters, Jan ; Georgia Institute of Technology. Institute for Robotics and Intelligent Machines ; Technische Universität Darmstadt. Computer Science Department ; Max-Planck Institute for Intelligent Systems
    Autonomous robots that can assist humans in situations of daily life have been a long standing vision of robotics, artificial intelligence, and cognitive sciences. A first step towards this goal is to create robots that can learn tasks triggered by environmental context or higher level instruction. However, learning techniques have yet to live up to this promise as only few methods manage to scale to high-dimensional manipulator or humanoid robots. In this talk, we investigate a general framework suitable for learning motor skills in robotics which is based on the principles behind many analytical robotics approaches. It involves generating a representation of motor skills by parameterized motor primitive policies acting as building blocks of movement generation, and a learned task execution module that transforms these movements into motor commands. We discuss learning on three different levels of abstraction, i.e., learning for accurate control is needed to execute, learning of motor primitives is needed to acquire simple movements, and learning of the task-dependent “hyperparameters“ of these motor primitives allows learning complex tasks. We discuss task-appropriate learning approaches for imitation learning, model learning and reinforcement learning for robots with many degrees of freedom. Empirical evaluations on a several robot systems illustrate the effectiveness and applicability to learning control on an anthropomorphic robot arm. These robot motor skills range from toy examples (e.g., paddling a ball, ball-in-a-cup) to playing robot table tennis against a human being and manipulation of various objects.
  • Item
    Robot Motion Planning: Challenges and Opportunities for Increasing Robot Autonomy
    (Georgia Institute of Technology, 2022-01-26) Moll, Mark ; Georgia Institute of Technology. Institute for Robotics and Intelligent Machines ; PickNik
    Robot manipulators are increasingly deployed outside of carefully controlled factory settings. Advances in robot motion planning have made it possible to compute feasible motions for more complex systems. My work is focused on enabling planning over varying time horizons subject to complex soft and hard constraints. The goal is to reduce the amount of user input required to command a robot and enable ever greater levels of autonomy. In this presentation I will first give a brief overview of sampling-based motion planning, a class of methods that has been successfully applied to a broad range of complex systems. I will present recent results that show that satisfying hard constraints can be decoupled from the particular planning strategy, which can lead to surprising performance improvements. Next, I will present some results on using hyperparameter optimization to select and tune motion planning algorithms for a given robot. Finally, will present some initial results on supervised autonomy that combines motion planning with compliant control, perception, and human input.
  • Item
    Robust Control Tools for Validating UAS Flight Controllers
    (Georgia Institute of Technology, 2022-11-16) Farhood, Mazen ; Georgia Institute of Technology. Institute for Robotics and Intelligent Machines ; Virginia Polytechnic Institute and State University. Kevin T. Crofton Department of Aerospace and Ocean Engineering
    This talk presents a framework based on robust control theory to aid in the certification process of unmanned aircraft system (UAS) flight controllers. Uncertainties are characterized and quantified based on mathematical models and flight test data obtained in-house for a small, commercial, off-the-shelf platform with a custom autopilot. These uncertainties are incorporated via a linear fractional transformation to model the uncertain UAS. Utilizing integral quadratic constraint (IQC) theory to assess the uncertain UAS worst-case performance, it is demonstrated that this framework can determine system sensitivities to uncertainties, compare the robustness of controllers, tune controllers, and indicate when controllers are not sufficiently robust. To ensure repeatability, this framework is used to tune, compare, and analyze a suite of controllers, including path-following, trajectory-tracking, H-infinity, H2, and PID controllers. By employing a non-deterministic simulation environment and conducting numerous flight tests, it is shown that the uncertain UAS framework reliably predicts loss of control, compares the robustness of different controllers, and provides tuned controllers which are sufficiently robust. Furthermore, robust performance guarantees from IQC analysis can be used to provide worst-case bounds on the UAS state at each point in time, providing an inexpensive and robust mathematical tool to aid in the certification of UAS flight controllers.