By Sara Downey, Dell Technologies
When the coronavirus spread through China, government officials called on its tech sector to help contain the virus. China was already a heavy investor in robotics—the robotics market was poised to grow from 26.9 percent of the world’s total spend on robots in 2017 to 38.3 percent by 2022, according to IDC. So when the need arose, the country was well-prepared to transition its growing battalion of machines to COVID-19-related duties.
One such robot included Little Peanut, which roamed the halls of a Hangzhou hotel to deliver food to patients. Meanwhile, 5G-powered robots were dispatched to a Shanghai hospital to support consultations, disinfection, cleaning, and drug delivery. Many machines, originally intended for other purposes, were redeployed to serve on the front lines. Shenzhen-based Pudu Technology, which normally makes robots for the catering industry, for instance, installed its machines in more than 40 hospitals around the country to aid medical staff.
Like China, other countries have called on robots to tend to patients from a safe distance. The first person diagnosed in the United States was treated by a robot equipped with a stethoscope. Globally, however, instances of robot intervention are still novel. According to rescue-robot trailblazer Dr. Robin Murphy, professor of computer science and engineering at Texas A&M University, these robots are too few and far between to fend off today’s pandemic. Which begs the question, how can we learn from the struggles today and progress global research efforts to develop robots at scale, so we’re better prepared for challenges in the future?
Going Where No Human Should
There is growing awareness among governments and aid agencies that the way we respond to a crisis doesn’t have to be curtailed by mere mortal capabilities: We can harness technology to do the seemingly impossible in dangerous terrains and under adverse conditions.
In fact, use cases for robots in disaster zones abound—from fighting fires and delivering medical supplies to carrying out reconnaissance work in highly radioactive environments. Take Little Sunfish, a robot the size of a loaf of bread, that ventured into the Fukushima Daiichi nuclear power plant in July 2017. Donning sensors, lights, and powerful propellers, the machine successfully identified the location of melted uranium fuel in one of Fukushima’s condemned reactors, bringing engineers a step closer to removing the deadly fuel and stemming contamination.
Dr. Murphy equates robots’ raison d’être to that of a fire truck: “They can do things that we can’t—dangerous, life-saving things.”
In this crisis-prone world, Dr. Murphy is looking to Germany as an emerging leader in disaster robotics. She’s particularly enthusiastic about the German Rescue Robots Center (DRZ) in Dortmund, which is prototyping intelligent robots.
“These robots are vastly different to industrial robots or analogue mobile robots like those of 1986, which were planted on the roof of a nuclear plant in Chernobyl to clean up parts of the destroyed reactor,” says Robert Grafe, managing director of DRZ. “Today, we’re developing robots with semi-autonomous abilities to gather and process relevant information, suppress fires, check measurements, read valves, climb stairs, and open doors.”
“Today, we’re developing robots with semi-autonomous abilities to gather and process relevant information, suppress fires, check measurements, read valves, climb stairs, and open doors.”
—Robert Grafe, managing director, German Rescue Robots Center
Another German hub of expertise is the DFKI GmbH Robotics Innovation Center (RIC) in Bremen. Dr. Florian Cordes, team leader of the terrestrial robotics group at RIC, emphasizes the importance of field testing. He points to the group’s four-wheel drive robot SherpaTT, which underwent eight weeks of testing: four weeks in the desert of Utah alongside RIC’s fully integrated rover system Coyote III, and four weeks in the desert of Morocco. While both SherpaTT and Coyote III are built for space exploration, they can also be adapted for life-saving missions here on Earth.
Dr. Cordes is very passionate about using these robots for search and rescue and foresees them being used amongst collapsed buildings—carrying heavy equipment to free up the rescuer to focus on the mission—or detecting hazardous chemicals and making the ground safe. However, he warns that they’ll be liability if they’re not rigorously tested in real-life scenarios. “The highest gap to bridge is the reliability of the solutions. They might work once in lab but they have to work all the time in the field. In search and rescue scenarios, there is no option to fail.”
“They [robots] might work once in lab but they have to work all the time in the field. In search and rescue scenarios, there is no option to fail.”
—Dr. Florian Cordes, team lead, DFKI GmbH Robotics Innovation Center
Dr. Cordes notes that some of this risk can be mitigated by humans partnering with machines: “Of course, humans will still be at the helm, but they’ll work at a safe distance.”
For the time being, the primary use of disaster robots will be to see and act remotely—and then feed information back, so people can make life-saving decisions in the moment. However, Dr. Murphy cautions that it’s extremely difficult to program for the unknown. “Disaster situations vary hugely, are unpredictable, and change quickly. Robots would have to be able to collect a vast amount of data and then apply protocols to it, so the right amount and type of data can be relayed and acted upon.”
“They should really be called ‘data robots,’ as their true value lies in the data they can collect and communicate. This is where the bleeding-edge innovation is happening.”
—Dr. Robin Murphy, professor of computer science and engineering, Texas A&M University
As we step into the data decade, Dr. Murphy believes the term “disaster robots” will become a misnomer. “They should really be called ‘data robots,’ as their true value lies in the data they can collect and communicate. This is where the bleeding-edge innovation is happening.”
From Lab to Land
Transitioning innovations from the laboratory to the real world can be fraught with difficulty, particularly for disaster robots used in life-or-death situations. According to Grafe, they need to meet tough criteria: “They must be incredibly robust, simple to use in a mechanical world, and, of course, fail-safe. But they also need to reassure users that they’re there to make their work safer rather than replace them. How they collaborate with humans is key.”
Grafe observes that commercializing robotics can equally be a challenge: “The specialist hardware is very expensive and difficult to source. What’s more, the public’s resistance to try new things needs to be overcome.”
However, Dr. Christian Illing of the German Federal Agency for Technical Relief (THW) believes that with sufficient access to demonstrators to test the machines onsite, people can be persuaded. He sees huge advantages in embedding these demonstrators early on: “We’ve applied for funding to develop intelligent systems that can respond to gestures, like the wave of a hand and do 3D printing on site. But we’ve also changed our funding applications to take our research a step further. We want to place our prototypes in the hands of firefighters for instance, gather feedback and develop further.”
As governments across the globe look for ways to tackle COVID-19, Dr. Murphy comes back to the intersection between robot and human. “It may be too late to partner with robots to treat large volumes of infected patients today and protect our emergency services, but we can use this groundswell of activity to prepare for future scenarios.”
Dr. Cordes agrees. He believes that if we overcome the barriers to transitioning from prototype to large-scale deployment, we’ll find robots indispensable. “Once they’re here they will be used—just like how we use the car rather than the horse-drawn carriage,” he says. “One day we might actually struggle to imagine life without them.”