From self-driving cars, to drones, to healthcare robotics: UC San Diego at ICRA 2021 preview
April 26, 2021-- A record number of papers from robotics faculty at the University of California San Diego were accepted to the 2021 International Conference on Robotics and Automation taking place in Xi’an.China, May 30 to June 5.
From helping robots navigate the ER, to making it easier for autonomous drones to fly around obstacles, to teaching robots how to suture wounds, the papers demonstrate the breath and depths of robotics research taking place at the UC San Diego Contextual Robotics Institute.
“It is encouraging to see the strong growth in submissions in all areas of robotics research,” said Henrik Christensen, director of the robotics institute, which includes more than 50 faculty and more than 100 graduate students. “In spite of COVID-19 and all the challenges it presents, our research work continues in full force.”
Below is a list of papers accepted to the conference, in alphabetical order by department, with summaries and links.
Department of Computer Science and Engineering
Looking Farther in Parametric Scene Parsing with Ground and Aerial Imagery
Raghava Modhugu, Harish Rithish Sethuram, Manmohan Chandraker, C.V. Jawahar
In this paper, researchers demonstrate the effectiveness of using aerial imagery as an additional modality to overcome challenges in road scene understanding. We propose a novel architecture, Unified, that combines features from both aerial and ground imagery to infer scene attributes.
Auto-calibration Method Using Stop Signs for Urban Autonomous Driving Applications
Yunhai Han, Yuhan Liu, David Paz, Henrik Christensen
Researchers developed an approach to calibrate sensors for self-driving cars using recognition of traffic signs, such as stop signs. The approach is based on detection, geometry estimation, calibration, and recursive updating. Results from natural environments are presented that clearly show convergence and improved performance.
Social Navigation for Mobile Robots in the Emergency Department
Angelique Taylor, Sachiko Mastumoto, Wesley Xiao, and Laurel D. Riek
In this paper, we introduce the Safety-Critical Deep Q-Network (SafeDQN) system, a new acuity-aware navigation system for mobile robots.
Temporal Anticipation and Adaptation Methods for Fluent Human-Robot Teaming
Tariq Iqbal and Laurel D. Riek
In this paper, we introduce TANDEM: Temporal Anticipa- tion and Adaptation for Machines, a series of neurobiologically- inspired algorithms that enable robots to fluently coordinate with people. TANDEM leverages a human-like understanding of external and internal temporal changes to facilitate coordination.
Department of Electrical and Computer Engineering
Mesh Reconstruction from Aerial Images for Outdoor Terrain Mapping Using Joint 2D-3D Learning
Q. Feng, N. Atanasov
This paper develops a joint 2D-3D learning approach to reconstruct local meshes at each camera keyframe, which can be assembled into a global environment model. Each local mesh is initialized from sparse depth measurements.
Coding for Distributed Multi-Agent Reinforcement Learning
B. Wang, J. Xie, N. Atanasov
We propose a coded distributed learning framework, which speeds up the training of MARL algorithms in the presence of stragglers, while maintaining the same accuracy as the centralized approach. As an illustration, a coded distributed version of the multi-agent deep deterministic policy gradient(MADDPG) algorithm is developed and evaluated
Non-Monotone Energy-Aware Information Gathering for Heterogeneous Robot Teams
X. Cai, B. Schlotfeldt, K. Khosoussi, N. Atanasov, G. J. Pappas, J. How
This work proposes a distributed planning approach based on local search, and shows how to reduce its computation and communication requirements without sacrificing algorithm performance.
Active Bayesian Multi-class Mapping from Range and Semantic Segmentation Observations
A. Asgharivaskasi, N. Atanasov
This work develops a Bayesian multi-class mapping algorithm utilizing range-category measurements. We derive a closed-form efficiently computable lower bound for the Shannon mutual information between the multi-class map and the measurements.
Learning Barrier Functions with Memory for Robust Safe Navigation
K. Long, C. Qian, J. Cortes, N. Atanasov
This paper investigates safe navigation in unknown environments, using onboard range sensing to construct control barrier functions online. To represent different objects in the environment, we use the distance measurements to train neural network approximations of the signed distance functions incrementally with replay memory. This allows us to formulate a novel robust control barrier safety constraint which takes into account the error in the estimated distance fields and its gradient.
Generalization in reinforcement learning by soft data augmentation
Nicklas Hansen, Xiaolong Wang
Instead of learning policies directly from augmented data, we propose SOft Data Augmentation (SODA), a method that decouples augmentation from policy learning. Specifically, SODA imposes a soft constraint on the encoder that aims to maximize the mutual information between latent representations of augmented and non-augmented data, while the RL optimization process uses strictly non-augmented data.
Bimanual Regrasping for Suture Needles using Reinforcement Learning for Rapid Motion Planning
Z.Y. Chiu, F. Richter, E.K. Funk, R.K. Orosco, M.C. Yip
In this work, we present rapid trajectory generation for bimanual needle regrasping via reinforcement learning (RL). Demonstrations from a sampling-based motion planning algorithm is incorporated to speed up the learning. In addition, we propose the ego-centric state and action spaces for this bimanual planning problem, where the reference frames are on the end-effectors instead of some fixed frame.
Real-to-Sim Registration of Deformable Soft-Tissue with Position-Based Dynamics for Surgical Robot Autonomy
F. Liu, Z. Li, Y. Han, J. Lu, F. Richter, M.C. Yip
In this work, we propose an online, continuous, real-to-sim registration method to bridge from 3D visual perception to position-based dynamics(PBD) modeling of tissues. The PBD method is employed to simulate soft tissue dynamics as well as rigid tool interactions for model-based control.
Model-Predictive Control of Blood Suction for Surgical Hemostasis using Differentiable Fluid Simulations
J. Huang*, F. Liu*, F. Richter, M.C. Yip
We investigate the integration of differentiable fluid dynamics to optimizing a suction tool's trajectory to clear the surgical field from blood as fast as possible. The fully differentiable fluid dynamics is integrated with a novel suction model for effective model predictive control of the tool.
SuPer Deep: A Surgical Perception Framework for Robotic Tissue Manipulation using Deep Learning for Feature Extraction
J. Lu, A. Jayakumari, F. Richter, Y. Li, M.C. Yip
In this work, we overcome the challenge by exploiting deep learning methods for surgical perception. We integrated deep neural networks, capable of efficient feature extraction, into the tissue reconstruction and instrument pose estimation processes.
Data-driven Actuator Selection for Artificial Muscle-Powered Robots
T. Henderson, Y. Zhi, A. Liu, M.C. Yip
To accelerate the development of artificial muscle applications, researchers developed a data driven approach for robot muscle actuator selection using Support Vector Machines (SVM). This first-of-its-kind method gives users insight into which actuators fit their specific needs and actuation performance criteria, making it possible for researchers and engineers with little to no prior knowledge of artificial muscles to focus on application design. It also provides a platform to benchmark existing, new, or yet-to-be-discovered artificial muscle technologies.
Optimal Multi-Manipulator Arm Placement for Maximal Dexterity during Robotics Surgery
J. Di, M. Xu, N. Das, M.C. Yip
Engineers created a method to generate the optimal manipulator base positions for the multi-port da Vinci surgical system that minimizes self-collision and environment-collision, and maximizes the surgeon's reachability inside the patient. The metrics and optimization strategy are generalizable to other surgical robotic platforms so that patient-side manipulator positioning may be optimized and solved.
J. Di, M. Xu, N. Das, M.C. Yip
MPC-MPNet: Model-Predictive Motion Planning Networks for Fast, Near-Optimal Planning under Kinodynamic Constraints
L. Li, Y.L. Miao, A.H. Qureshi, M.C. Yip
We present a scalable, imitation learning-based, Model-Predictive Motion Planning Networks framework that quickly finds near-optimal path solutions with worst-case theoretical guarantees under kinodynamic constraints for practical underactuated systems. Our framework introduces two algorithms built on a neural generator, discriminator, and a parallelizable Model Predictive Controller (MPC).
F. Richter, S. Shen, F. Liu, J. Huang, E.K. Funk, R.K. Orosco, M.C. Yip
In this work, we present the first, automated solution for hemostasis through development of a novel probabilistic blood flow detection algorithm and a trajectory generation technique that guides autonomous suction tools towards pooling blood. The blood flow detection algorithm is tested in both simulated scenes and in a real-life trauma scenario involving a hemorrhage that occurred during thyroidectomy.
Department of Mechanical and Aerospace Engineering
Scalable Learning of Safety Guarantees for Autonomous Systems using Hamilton-Jacobi Reachability
Sylvia Herbert, Jason J. Choi, Suvansh Qazi, Marsalis Gibson, Koushil Sreenath, Claire J. Tomlin
In this paper we synthesize several techniques to speed up computation: decomposition, warm-starting, and adaptive grids. Using this new framework we can update safe sets by one or more orders of magnitude faster than prior work, making this technique practical for many realistic systems.
Planning under non-rational perception of uncertain spatial costs
Aamodh Suresh and Sonia Martinez
This work investigates the design of risk- perception-aware motion-planning strategies that incorporate non-rational perception of risks associated with uncertain spatial costs. Our proposed method employs the Cumulative Prospect Theory (CPT) to generate a perceived risk map over a given environment.
Jacobs School of Engineering