Morphy: a Compliant and Morphologically-Aware Flying Robot
We present Morphy*, a novel Compliant and Morphologically-aware Flying Robot that integrates sensorized flexible joints in its arms, thus enabling resilient collisions at high speeds and the ability to squeeze through openings more narrow than its nominal dimensions. The sensorized soft joints include 3D Hall-effect sensors therefore delivering real-time updates for the robot's adapting morphology, especially during collisions and other instances of physical interaction. Morphy weighs only 260 g including its battery, camera, time-of-flight depth sensing and advanced 8-core processor meaning that it is "autonomy-capable". The complete frame is built through adaptive manufacturing techniques, including elastic resin printing.
AI for Good: Robotic autonomy in the wild
Prof. Alexis, SPEAR Coordinator, provided a talk at the AI for Good Global Summit 2024 organized by ITU of the United Nations. In this presentation, Prof. Alexis outlined a vision for resilient autonomy, provided live demonstration of robotic systems together with members of his lab and finally pointed towards a superior technological feature especially through the research direction of SPEAR. Computational co-design of aerial robot frames and intelligence shall lead to systems with unprecedented skills and capacities!
Vine Robots that Evert through Bending
Despite the elegantly simple operation principle, the design of everting vine robots still presents two complexities: a tip-device required for buckle-free retraction; a pressurised base chamber to maintain inflation while allowing the everted body to be winded onto a motorised reel. We create a new type of vine robots with a unique eversion method: instead of using a single everting tube, our design comprises multiple interconnected tubes that bend to induce an overall eversion. Our design eliminates the pressurised base and the tip-device, while allowing buckle-free retraction. It also enables new designs and functionalities beyond conventional vine robots, including drone-based vine robots capable of cantilevered operations, everting grippers that grasp then transport objects into the robot body, and solid-state vine robots without the need for inflation.
A Terminal State Feasibility Governor for Real-Time Nonlinear MPC over Arbitrary Horizons
This article introduces a novel feasibility governor (FG), which enlarges the region of attraction (ROA) of a nonlinear model predictive control (NMPC) setpoint regulation law with an arbitrarily short prediction horizon. The efficient online FG is developed for nonlinear systems subject to pointwise-in-time state and input constraints and relies on a discrete-time solution of a trajectory-based explicit reference governor (ERG) with Lyapunov-based terminal energy constraint.
D-MARL: A Dynamic Communication-Based Action Space Enhancement for Multi Agent Reinforcement Learning Exploration of Large Scale Unknown Environments
In this article, we propose a novel communication-based action space enhancement for the D-MARL exploration algorithm to improve the efficiency of mapping an unknown environment, represented by an occupancy grid map.
Belief Scene Graphs: Expanding Partial Scenes with Object through Computation of Expectation
In this article, we propose the novel concept of Belief Scene Graphs, which are utility-driven extensions of partial 3D scene graphs, that enable efficient high-level task planning with partial information. We propose a graph-based learning methodology for the computation of belief (also referred to as expectation) on any given 3D scene graph, which is then used to strategically add new nodes (referred to as blind nodes) that are relevant to a robotic mission.
Leveraging Computation of Expectation Models for Commonsense Affordance Estimation on 3D Scene Graphs
This article studies the commonsense object affordance concept for enabling close-to-human task planning and task optimization of embodied robotic agents in urban environments. The focus of the object affordance is on reasoning how to effectively identify object's inherent utility during the task execution, which in this work is enabled through the analysis of contextual relations of sparse information of 3D scene graphs.
Novel Efficient RLSimulation Benchmark for End-to-End Vision-Only Unmanned Aerial Vehicle Navigation
In this work, we developed a simulation-in-the-loop using NVIDIA ISAAC SIM to train navigation policies in an end-to-end manner. Like a human, the agent makes decisions based solely on visual sensory inputs from an RGB-D camera. We trained the agent in two different scenarios: (a) a warehouse with static obstacles and (b) a collapsed building environment with dynamic obstacles. The results demonstrate that the agent can safely navigate in extreme settings.
Aerial Gym Simulator: A Framework for Highly Parallelized Simulation of Aerial Robots
We are happy to release an update for the Aerial Gym Simulator. The new update includes support for multi-linked embodiments, with fixed, reconfigurable (active), and soft (passive) joints for simulated multirotors. A faithful simulation model of a compliant robot Morphy and models for active reconfigurable multirotor platforms are added!
The Unified Autonomy Stack: Toward a Blueprint for Generalizable Robot Autonomy
We introduce and open-source the Unified Autonomy Stack, a system-level solution that enables resilient autonomy across diverse aerial and ground robot morphologies. The architecture centers on three synergistic modules - multi-modal perception, multi-behavior planning, and multi-layered safe navigation - that together deliver comprehensive mission autonomy. T
Neural NMPC through Signed Distance Field Encoding for Collision Avoidance
This paper introduces a Neural Nonlinear Model Predictive Control Framework (NMPC) framework for mapless, collision-free navigation in unknown environments with Aerial Robots, using onboard range sensing. We leverage deep neural networks to encode a single range image, capturing all the available information about the environment, into a Signed Distance Field (SDF).
Performance-guided Task-specific Optimization for Multirotor Design
We introduce a methodology for task-specific design optimization of multirotor Micro Aerial Vehicles. By leveraging reinforcement learning, Bayesian optimization, and covariance matrix adaptation evolution strategy, we optimize aerial robot designs guided only by their closed-loop performance in a considered task.
Circle CPPN Best Non-Speciated
Results in evolved drone morphologies. Associated playlist.
UniPilot: Enabling GPS-Denied Autonomy across Embodiments
This work presents UniPilot, a compact hardware-software autonomy payload that can be integrated across diverse robot embodiments to enable resilient autonomous operation in GPS-denied environments. The system integrates a multi-modal sensing suite including LiDAR, radar, vision, and inertial sensing for robust operation in conditions where uni-modal approaches may fail.
Embodiment-conditioned Generalist Control for Multirotor Aerial Robots
We present a generalist position control policy capable of controlling arbitrary multirotor configurations of a certain rotor count (e.g., hexarotors or quadrotors) with a single set of network weights. The policy is conditioned on a physics-grounded embodiment descriptor: a mass and inertia-normalized control allocation matrix that captures how mass-normalized motor thrusts generate linear and angular accelerations in the body-frame.