Telerobotic Operation of Intensive Care Unit Ventilators

Balazs P. Vagvolgyi; Mikhail Khrenov; Jonathan Cope; Anton Deguet; Peter Kazanzides; Sajid Manzoor; Russell H. Taylor; Axel Krieger

arXiv · 2010.05247

Telerobotic Operation of Intensive Care Unit Ventilators

Balazs P. Vagvolgyi·Mikhail Khrenov·Jonathan Cope·Anton Deguet·Peter Kazanzides·Sajid Manzoor·Russell H. Taylor·Axel Krieger

View original source ↗Read source document ↗Export metadata ↓

Abstract

Since the first reports of a novel coronavirus (SARS-CoV-2) in December 2019, over 33 million people have been infected worldwide and approximately 1 million people worldwide have died from the disease caused by this virus, COVID-19. In the US alone, there have been approximately 7 million cases and over 200,000 deaths. This outbreak has placed an enormous strain on healthcare systems and workers. Severe cases require hospital care, and 8.5\% of patients require mechanical ventilation in an intensive care unit (ICU). One major challenge is the necessity for clinical care personnel to don and doff cumbersome personal protective equipment (PPE) in order to enter an ICU unit to make simple adjustments to ventilator settings. Although future ventilators and other ICU equipment may be controllable remotely through computer networks, the enormous installed base of existing ventilators do not have this capability. This paper reports the development of a simple, low cost telerobotic system that permits adjustment of ventilator settings from outside the ICU. The system consists of a small Cartesian robot capable of operating a ventilator touch screen with camera vision control via a wirelessly connected tablet master device located outside the room. Engineering system tests demonstrated that the open-loop mechanical repeatability of the device was 7.5\,mm, and that the average positioning error of the robotic finger under visual servoing control was 5.94\,mm. Successful usability tests in a simulated ICU environment were carried out and are reported. In addition to enabling a significant reduction in PPE consumption, the prototype system has been shown in a preliminary evaluation to significantly reduce the total time required for a respiratory therapist to perform typical setting adjustments on a commercial ventilator, including donning and doffing PPE, from 271 seconds to 109 seconds.

Explore related subjects

cs.RO

Keep this discovery

BibTeX RIS

Balazs P. Vagvolgyi, Mikhail Khrenov, Jonathan Cope, Anton Deguet, Peter Kazanzides, Sajid Manzoor, Russell H. Taylor, Axel Krieger. 2020-10-11. Telerobotic Operation of Intensive Care Unit Ventilators. https://arxiv.org/abs/2010.05247

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

IMLE-VLA: Fast Single-Step Action Generation for Vision-Language-Action Policies

Vision-language-action (VLA) policies leverage pretrained vision-language backbones to achieve strong cross-task generalization. A leading design couples this backbone with a dedicated continuous action head trained via diffusion or flow matching. However, such heads rely on iterative multi-step sampling, for example 10 Euler steps in $\pi_{0.5}$. This creates an inference bottleneck that produces stop-and-go movement in the robot and slower task completion. We introduce IMLE-VLA, which replaces the iterative action head with a single-step conditional generator trained via conditional Implicit Maximum Likelihood Estimation (cIMLE). The cIMLE objective promotes multimodal action coverage, avoiding the mode collapse of naive regression heads while eliminating multi-step sampling entirely. When IMLE-VLA is applied to $\pi_{0.5}$, it increases inference frequency 3.67x (55 Hz vs. 15 Hz), enabling up to 11x higher action throughput. On the 40-task LIBERO benchmark, IMLE-VLA achieves the highest average success rate (98.0%) among all baselines while leading in inference frequency. Under the test-time perturbations of LIBERO-plus, IMLE-VLA retains $\pi_{0.5}$'s robustness while other baselines degrade sharply, confirming that the cIMLE head preserves generalization. Real-world experiments on a Franka Emika Panda across four tasks demonstrate smoother motion (2.2x to 3.0x lower jerk) and faster task completion, with IMLE-VLA outperforming $\pi_{0.5}$ on every task and reducing average VLA inference time per episode by 3.9x to 6.6x. Videos and code are available at https://kianhk6.github.io/IMLE-VLA/

cs.RO↗

arXiv2026

ObstaDiff: Generalizable Diffusion Policy Learning via Obstacle-aware Representations

Imitation learning has achieved impressive results in robotic manipulation, yet most existing approaches assume clean backgrounds and lack explicit mechanisms for obstacle-aware motion generation. Extending such policies to cluttered, real-world scenes with unstructured obstacles remains a key generalization challenge. We present ObstaDiff, a decomposed diffusion-policy framework with a lightweight obstacle-aware visual encoder. ObstaDiff extracts a structured target-obstacle-background representation, enabling the downstream alignment policy to generate end-effector trajectories toward a target-centered bottleneck pose while reasoning about surrounding obstacles. We evaluate ObstaDiff on 61 real-robot greenhouse trials per method (366 executions in total). ObstaDiff achieves 75.41% average task success and 8.20% average obstacle collision rate, outperforming representative imitation-learning baselines and improving generalization in cluttered agricultural scenes.

cs.RO↗

arXiv2026

Testing Between the Test Cases: Proving End-to-End Steering in Conditions You Never Drove

AI-based automated vehicle testing is challenging because a model that passes every test condition can still fail in the real world. Formal verification offers a way to directly address this gap. On a simulated highway and an arterial road we trained two small end-to-end steering networks each in CARLA, one on clear conditions alone and one on clear, fog, night and low sun. All four models were driven against a 2.19 ft lane-departure budget. Without driving again, we used bound propagation, a formal method that reads the trained weights, to compute how far steering can drift at every disturbance strength between two captured images. One calculation covers more than a campaign could drive: on the arterial it spans 133 poses, where ten intensities each would be 10^133 combinations, in minutes on one GPU. Not only did formal verification find conditions that broke the clear-trained policy without simulation testing, it provided some preliminary evidence for potential failures between the test cases. Our overall conclusion is that formal verification is a viable complement to simulation, and could be adopted as a part of verification and validation for automated driving.

cs.RO↗