Scientists consider repurposing robots to contain Ebola epidemic

Robotics scientists nationwide are pondering an intriguing possibility: Might robotic technologies deployed in rescue and disaster situations be quickly repurposed to help contain the Ebola epidemic?

A robot that could perform even some of the tasks of a human, such as waste removal or the burial of bodies, would have significant lifesaving potential. So, with the assistance of the White House Office of Science and Technology Policy, scientists are planning a series of brainstorming meetings. The first round will be held on 7 November at four locations: Worcester Polytechnic Institute, in Massachusetts; Texas A&M; the University of California, Berkeley; and in Washington.

The problem, scientists say, is that the technology is still limited when it comes to medicine. While mobile robots now can disarm roadside bombs and drive cars, they are taking only the first tentative steps toward the human levels of dexterity required in health care.

"You see the situation that the medical teams are facing, and I don't even know if a robot is a solution," said Taskin Padir, an assistant professor of robotics engineering at Worcester Polytechnic and an organiser of one of the meetings.

Still, he has been considering ways to repurpose an existing robot project as a tool for more safely performing decontamination tasks, such as spraying bleach solution on clothing exposed to infected body fluids.

"As was the case in Fukushima, the Ebola crisis in Africa has revealed a significant gap between robot capabilities and what is needed in the realm of disaster relief and humanitarian assistance," said Gill A Pratt, a roboticist who is a program manager at the federal Defense Advanced Research Projects Agency. "We have a moral obligation to try and select, adapt and apply available technology where it can help, but we must also appreciate the difficulty of the problem."

Many of the countries experiencing the worst of the Ebola epidemic are in no position to deploy robots. But rudimentary models are widely used in medical settings in the United States, and already nervous hospital administrators are reaching out to manufacturers.

The phone has been ringing continually in recent weeks, said Yulun Wang, the chief executive of InTouch Health, a Santa Barbara, California-based maker of telepresence robots used in hospitals for stroke diagnosis and other medical tasks.

Clients such as Robert Wood Johnson University Hospital and Baylor University Hospital are asking whether the robots can be used to help diagnose Ebola infection, without a human present, or to facilitate virtual family visits for patients in isolation.

"They acquired our solutions for a very different purpose, and now they are wondering whether they are applicable for Ebola care," Wang said.

The company makes a robot that travels around hospital corridors, guided by a physician at a remote location who "sees" via an Internet-connected device. Normally used to observe patients with the aid of high-quality video and audio, the robots could also be programmed to train health care personnel, Wang said.

"There isn't a lot of Ebola expertise in the United States," he said. "We can beam those experts into the 1,000 hospitals that have our systems instantly." Robots might be programmed to provide interactive checklists to medical personnel who are putting on and taking off gowns, much in the way aircraft pilots work from checklists on takeoff and landing. Some experts have also suggested that telepresence robots might help language interpreters offer assistance in the field.

But dexterous robots that move without human guidance are only beginning to appear in laboratories. A robot that could work in close contact with a patient, replacing a nurse or a doctor, is still years away.

Humanoid rescue robots were on display in December in Florida at a contest created by Darpa, part of the Defense Department, which is hoping to spark innovation in the design of machines that are capable of operating autonomously in hostile environments.

But even the best robot, designed by a group of Japanese researchers, could drive a car only in a halting fashion, and only after the robot was attached to the steering wheel and brake and accelerator pedals. The challenges in a health care setting could be even greater.

One small example: Glass beakers and test tubes are extremely challenging for machine vision systems. Even the newest and least expensive versions, like Microsoft's Kinect, struggle to recognise transparent glass objects.

Research efforts begun now may not bear fruit until the next epidemic, according to a number of the researchers.

Ken Goldberg, a professor of industrial engineering and operations research at the University of California, Berkeley, worries that the robotics researchers might seem opportunistic if they rush forward with unproven technology.

"We don't want to be seen as capitalising on the tragedy," Dr. Goldberg said. "You don't want to be seen as, 'We're sending in the robots.' It sounds insensitive and crass."

Scientists also cited cultural obstacles to the adoption of robots for medical uses.

"One of the first things I heard from medical responders is that one of the bottlenecks is in handling bodies," said Robin Murphy, a specialist in rescue robotics and a professor of computer science and engineering at Texas A&M.

Families often cannot accept the idea that a loved one's body might be handled by a machine. "It's something we can do, but it has to be culturally sensitive," she said.

©2014 The New York Times News Service