Hospital uses robots to manage air supply to patients

Northampton General Hospital is using automation software to monitor and manage oxygen supplies following a project spurred by the Covid-19 crisis.

Like other hospitals across the country, the coronavirus pandemic has seen Northampton General face a sudden rise in the number of patients requiring oxygen.

Managing oxygen flow had become a life-threatening problem at neighbouring hospitals, which prompted a project at Northampton General to apply robot process automation to the manual task of checking oxygen flow.

At the outbreak’s peak, the hospital was treating about 150 Covid-19 patients simultaneously, and that number has hovered around 100. These are severe cases as there are many people with Covid-19 who don’t go to hospital.

Because of the increased demand for hospital treatment during the pandemic, Northampton General has been carefully managed, including clearing space for hundreds of extra beds and building extra wards in theatre areas to help cope with the increased number of people needing ventilation, said its CIO, Hugo Mathias.

The hospital also instigated its first robotic process automation (RPA) project to make sure critical oxygen supplies were not interrupted.

“The problem was that other hospitals started getting problems around the capacity of oxygen,” said Mathias. “When one of our neighbouring hospitals got into trouble with the volume of oxygen it needed, which exceeded the volume of their tank, we were told to keep an eye on our oxygen levels.”

Northampton General has two large oxygen tanks that supply ventilation machines, and patients are assigned to hospital wards to allow for the equal distribution of oxygen. Monitoring the tanks would normally involve manually checking the oxygen levels every hour, which is challenging at night and something staff need to be reminded to do.

“It seemed silly to do something like this when we have all the information available online,” said Mathias. “A robot can get this information automatically.”

Mathias added: “The idea was to extract the data from the website that contains the tank information and put it on our servers, so we can do a calculation to work out the difference between readings. This tells us when we are getting close to maximum flow.

“We then collect information from the wards about who is using oxygen and how it was being delivered. This helped us load-balance oxygen demand across the wards.”

The biggest challenge, other than the high entry price, was “getting the skills to maximise the impact of something like RPA”, said Mathias. The 800-bed hospital trust has about 6,000 staff, with some 100 IT employees complemented by a group of medical record staff.

“The RPA that we wanted to do was challenging because, although we are a big trust, having the skills to put that into place is something that is foreign to us, so we didn’t have the skills to take on something new,” said Mathias.

However, the hospital did have the advantage of a non-executive director who is an expert in RPA. “He helped us get over that learning curve of what we had to do,” said Mathias. He also brokered a deal with RPA software supplier Automation Anywhere, he added.

“The project was easy because we had help from Automation Anywhere and its partner SBI,” said Mathias. “But we were still surprised just how easy it was. The project took a few days to complete, mainly because there were challenges in getting the software onto our system. As regards building the robot, it took around 10 hours, if that.”

It is too early to say how much time has been saved, but Mathias said he wants to use RPA to take away monotonous tasks and speed up some processes. “We have only had it a couple of weeks and this is very much the start of our journey,” he added.

Next up on that journey, the hospital will look at automating the processes of onboarding and offboarding staff, as well as processes around referrals. “We can extract and check data and highlight the errors that could cause delays in patient referrals,” said Mathias.