CPR Robot History and Where ABB Arms Could Help

Chest compressions are hard physical work. Rescuers tire within a couple of minutes, and in a moving ambulance good compressions are difficult to keep up at all. A CPR robot, or mechanical chest compression device, takes that job over and presses the chest at a steady rate and depth for as long as it is needed. This post follows CPR from the 1960 work on closed chest compressions at Johns Hopkins, through the first compressor machines from Grand Rapids, to the LUCAS and AutoPulse devices in ambulances now and the robots researchers are testing. It also looks at where ABB robots could fit, which for now is a possibility.

An automatic chest compression device fitted to a manikin, part of a Japan Air Self-Defense Force aeromedical unit, Komaki Air Base, February 2014. Photo: Hunini / Wikimedia Commons (CC BY-SA 4.0)

How rescuers kept hearts going by hand

Modern CPR began with a finding at Johns Hopkins. William Kouwenhoven, James Jude and Guy Knickerbocker showed that pressing hard on the chest, without opening it, could keep blood moving. The work was presented at the annual Maryland Medical Society meeting in Ocean City on 16 September 1960, alongside the mouth to mouth breathing that Peter Safar and James Elam had shown to work. Jude, Knickerbocker and Safar combined the two in a 1962 training film, The Pulse of Life, and in the United States CPR was taught to the public from the 1970s. Every compression was done by hand.

US Army medical personnel practise CPR by hand on a manikin in 1977. Photo: U.S. Army via Wikimedia Commons (public domain)

Engineers tried to automate the work almost at once. In 1962 Clare Barkalow, an engineer in the Advanced Engineering Projects department at Lear Siegler in Grand Rapids, Michigan, developed an External Cardiac Compressor. Lear Siegler was not interested, so in 1963 he founded Michigan Instruments, and by 1965 its Model 1001 for human use was on sale, a pneumatically powered compressor that sold about 100 units in the US. In 1968 the company combined compressor and ventilator in the Model 1003, Life-Aid CPR, the first hands-free system to do both. Dixie USA sold it as the Thumper.

Trainees take turns with compressions and a defibrillator on a Laerdal manikin, 2008. Photo: Ch-info.ch / Wikimedia Commons (CC BY 3.0)

The devices most crews carry today arrived around the turn of the century. LUCAS, short for Lund University Cardiopulmonary Assist System, came from Lund in Sweden and entered clinical use around 2002 to 2003. A backplate goes under the patient and a frame locks onto it, then an electrically driven piston pushes a suction cup pad on the breastbone, at factory settings of about 102 times a minute and about 53 mm deep. Early LUCAS units were pneumatic; later ones run on electric motors and batteries. ZOLL's AutoPulse squeezes the chest with a broad band instead, and the German Corpuls CPR uses a single cantilevered arm.

An AutoPulse band device on a manikin at the Rettmobil fair in Germany, May 2007. Photo: PhilippN / Wikimedia Commons (CC BY-SA 3.0)

CPR machines on the road in 2026

Fire and ambulance services keep buying them. In September 2026 the Meridianville Volunteer Fire Department in Madison County, Alabama, received a $42,000 grant from FEMA's Assistance to Firefighters Grant program for two LUCAS 3 devices, one on each primary response vehicle. The department said the machines give compressions that do not fatigue, free firefighters for other life saving work and keep patients safer during transport to hospital. Fire Chief Brandon Burgess called them "an extra set of tireless hands on the scene" that reduce the physical strain and safety risks for his crew.

An AutoPulse demonstration beside an ambulance at a nursing symposium at US Naval Hospital Okinawa, February 2025. Photo: Isaac Savitz, U.S. Navy via Wikimedia Commons (public domain)

Whether the machines save more lives is less clear. Guidelines ask for 100 to 120 compressions a minute, at least 5 cm deep, and by hand the depth starts to fall after about 90 to 120 seconds, which is why teams swap every two minutes. Yet in the PARAMEDIC trial of 4,471 patients, 30 day survival was 6.3 percent with LUCAS and 6.8 percent with manual CPR, and the LINC trial of 2,589 patients found 23.6 against 23.7 percent. New designs keep coming. At Seoul National University, Gil Joon Suh's team built ROSCER, a remote controlled compressor that can shift its compression point during CPR, and tested it against LUCAS 3 in pigs.

A Resusci Anne CPR manikin with its soft face cover off, the kind of dummy used for training and device tests. Photo: Pitke / Wikimedia Commons (CC BY-SA 4.0)

ABB makes no CPR devices, and I found no ABB robot used for CPR, so its fit is a possibility. A cobot such as the GoFa CRB 15000 has torque sensors in its joints and stops when it meets unexpected contact. That is the right reflex around people, and the opposite of what compressions need, which is firm and repeated pushing on a body. ABB arms are more likely to turn up around CPR than on the chest, for example testing compression devices on manikins hour after hour, or helping to build them, a field covered in our post on medical device manufacturing. ABB also runs a healthcare research hub where GoFa is used.

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

What the next CPR robots will do

Research is moving toward machines that respond to the patient. ROSCER points to robots that move the compression point instead of pressing one fixed spot. A 2025 paper in the Journal of Field Robotics described a soft robotic device with three pneumatic air muscles that compresses 80 to 120 times a minute at 260 N and performed on a CPR manikin about as well as manual compressions under AHA protocols. In 2024 the journal Resuscitation published work on an AI driven CPR robot that reacts to the patient's biosignals. For device makers, more sensing means more software to validate and a longer path to approval.

A US soldier gives compressions to a manikin during CPR training at Camp Bondsteel, Kosovo, February 2021. Photo: Sgt. Zachary Zippe, U.S. Army National Guard via Wikimedia Commons (public domain)

A CPR robot in the home is unlikely any time soon. The devices are heavy, need training to fit correctly and are expensive: Meridianville's $42,000 grant covered two of them. What reaches homes, offices and schools is the defibrillator and CPR training on manikins, some of which already measure the depth and rate of each push. For most people, the change will come through the ambulance or fire engine that arrives with a machine on board. Inside hospitals, robots are more likely to appear in the lab first, where ABB has shown a mobile YuMi concept for lab work.

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

The first open problem is evidence. Large trials have not shown that machines beat good manual CPR, so services that buy them point to transport and crew safety, as Meridianville did. A device has to be fitted fast, because every pause in compressions counts, and crews need practice to do that. A machine pressing on a person also needs tight limits on force and position, since compressions can break ribs or land in the wrong place. The force sensing and safety functions built into cobots like GoFa are one place such limits could come from, and any new design will need clinical trials before it rides in an ambulance.

Moving an ABB GoFa arm by hand; torque sensors in its joints feel the push and contact. Photo: ABB

Innovation

AI solutions for effortless ABB robot control.

Automation

Robotics

ceojohntran@chatwithrobot.net

+84905311611

© 2025. All rights reserved.

qtran1215@gmail.com