ICU Nursing Robot: Polio Wards to ABB Cobots

An ICU nursing robot is a machine that helps intensive care nurses with physical work at the bedside, such as getting a sedated patient moving again, fetching supplies or handling routine checks. The ICU itself is only about seventy years old, and for most of that time nurses did the exhausting physical work by hand while machines took over the breathing and the watching. This post looks at how the first intensive care units worked, what robots are doing in ICUs today, where ABB robots could fit, and what still stands in the way.

An intensive care bed with monitors and a stack of infusion pumps, 2025. Photo: Ismael Olea / Wikimedia Commons (CC BY 4.0)

How Copenhagen's polio wards created intensive care

Grouping the sickest patients together goes back at least to Florence Nightingale, who separated the seriously wounded from the rest during the Crimean War in 1854. The modern ICU came out of a crisis. In 1952 a polio epidemic hit Copenhagen, and in six months 2,722 people fell ill, 316 of them with paralysis that stopped them breathing on their own. The anaesthetist Bjørn Ibsen proposed keeping them alive with positive pressure ventilation through a tube in the windpipe. There were not enough machines, so about 200 medical students took turns squeezing air into patients' lungs by hand.

An iron lung ward full of polio patients at Rancho Los Amigos Hospital, California, around 1953. Photo: Food and Drug Administration via Wikimedia Commons (public domain)

Mortality among the paralysed patients fell from around 90 percent to about 25 percent. In 1953 Ibsen turned a former student nurse classroom at Copenhagen's Kommunehospitalet into what is usually called the world's first medical and surgical intensive care unit. Ibsen's point was that very sick patients do better when trained staff watch them constantly in one place. Hospitals in the United States opened similar units in the 1950s, and Swedish engineers built volume ventilators, such as the Engström respirator, that could take over the hand pumping the students had done.

An Engström respirator Model 150 from Sweden, used from 1955 to 1970. Photo: Science Museum Group / Wikimedia Commons (CC BY 4.0)

The next wave of automation was about watching rather than doing. In the 1960s doctors recognised heart rhythm problems as a major cause of death after heart attacks, and cardiac monitors became routine in intensive care. Bedside screens, alarms and central stations followed, so one nurse could keep an eye on several patients at once. Later, tele-ICU systems let specialists in another building follow vital signs and talk to the bedside team over video. None of these machines could turn a patient, carry a tray or help someone stand. In most units that work still falls to nurses.

A nurse tends a patient just out of surgery in the intensive care ward of a US Navy hospital ship, October 1967. Photo: US National Archives via Wikimedia Commons (public domain)

Robots at the ICU bedside in 2026

On 13 October 2026 Bioengineer.org reported on a concept analysis in BMC Nursing by Yingying Wang of the First Affiliated Hospital of Wannan Medical University and colleagues. The team searched nine English and Chinese databases up to 3 June 2026 and built a definition of robot-assisted care in critical care nursing from 14 reports. Their four attributes are a physical robot, sensing and task execution, a real place in ICU nursing workflows, and clinician control. Software that only predicts deterioration does not count, and neither does a robot that an engineering team runs without nursing oversight.

A nurse and a doctor at the bed of a ventilated patient in an intensive care unit. Photo: Calleamanecer / Wikimedia Commons (CC BY-SA 3.0)

The clearest use so far is early mobilisation. Weeks in bed cause ICU acquired weakness, and getting a patient up takes several staff; Bioengineer.org notes that each session can need two to four clinicians. VEMOTION, made by the Munich company Reactive Robotics, clamps onto the patient's bed, tilts it up to 70 degrees and moves the legs through knee adapters. In a randomised pilot at TUM University Hospital Rechts der Isar, published in October 2025, 18 neurocritical care patients took part. The robot group gained a median of 15 points on the CPAx function scale against 4 for controls, with no therapy related adverse events.

An intensive care room at AZ Sint-Maarten hospital in Mechelen, Belgium, with ceiling supply arms around the bed. Photo: Ad Meskens / Wikimedia Commons (CC BY-SA 4.0)

ABB does not make an ICU robot, and no ABB arm is in clinical use at the bedside. Its healthcare work so far is in laboratories: in October 2019 it opened a healthcare research hub at the Texas Medical Center in Houston and showed a mobile YuMi concept for hospital labs. The GoFa CRB 15000 cobot carries 5 kg, senses contact through torque sensors in its joints and can be guided by hand. It is possible to picture a GoFa sorting ICU supplies or loading sample racks near the unit, much like the hospital delivery logistics robots that already ride the lifts. ABB has announced nothing of the kind, so treat it as an idea.

A GoFa cobot at ABB's healthcare research hub. Photo: ABB

What ICU nursing robots still need to solve

Over the next few years, expect more robots that each do one physical job well. A general robot nurse is much further off. Mobilisation devices, patient turning and lifting aids, and mobile robots that bring drugs and linen to the unit are the obvious candidates, because the tasks are heavy, repetitive and easy to measure. Our post on patient lifting covers the lifting side. The BMC Nursing paper helps here. It asks hospitals to measure patient outcomes, nurse workload, workflow disruption, safety events and how patients feel about the machine, on top of checking that the device works.

A pharmacy delivery robot parked by a hospital nurse station, 2008. Photo: jk / Wikimedia Commons (CC BY 3.0)

Cost and evidence are the main brakes. Most studies are small, like the 18 patient VEMOTION pilot, and the authors of the BMC Nursing paper say preparation and coordination demands stayed high. Every device in an ICU must also be cleaned between patients, fit around ventilator tubes and lines, and stay safe when a patient suddenly deteriorates. For an industrial supplier like ABB, any move toward the bedside would mean medical device approval, which its lab automation work has not needed. Engineers can still test layouts and reach in software first, the way factory cells are planned in RobotStudio.

ABB's mobile YuMi concept, a dual-arm robot on a wheeled base, in a hospital lab corridor. Photo: ABB

The home is where many ICU survivors finish recovering, often with weak muscles and a long rehabilitation ahead. Robots that help with standing, walking and lifting at home are slowly appearing, and some ideas tested in intensive care will move there once they are cheaper and simpler. In factories the link runs the other way: the sensors, force limits and hand guiding that make cobots such as GoFa safe beside workers are the same features any bedside robot will need. Nurses will also need training to set up and supervise these machines, and the paper is clear that the nurse stays in charge.

Lead-through teaching: moving an ABB GoFa arm by hand. Photo: ABB

Innovation

AI solutions for effortless ABB robot control.

Automation

Robotics

ceojohntran@chatwithrobot.net

+84905311611

© 2025. All rights reserved.

qtran1215@gmail.com