Robots Break Free from the Factory--Part 2Larry Maloney | December 10, 2015
The world may well be entering the golden age of robotics. Figures from the International Federation of Robotics show that sales of industrial robots, such as those that weld panels on auto assembly lines, surged by nearly 30% in 2014 versus the previous year.
But there’s another class of robots that is also enjoying double-digit sales growth – so-called “field robots.” These increasingly sophisticated machines have taken robotics beyond the confines of the factory to perform tasks ranging from harvesting crops and mining coal to disarming bombs and performing delicate surgery.
In fact, the applications for such robots are limited only by our own imagination, says William “Red” Whittaker, who has spearheaded development of a long litany of field robots over the last 35 years and helped establish Pittsburgh’s Carnegie Mellon University as one of the world’s premier centers for robotics R&D. Field robots designed by Whittaker and his teams at CMU’s Robotics Institute have cleaned up nuclear waste, explored volcanoes, mapped mines and inspected sewer systems. Next up: launch of a lander and rover to the moon by Astrobotic Technology, a CMU spinoff.
Director of CMU’s Field Robotics Center and founder of his own robotics company, Whittaker shares his views on the potential of field robotics in an interview with Engineering360 contributing editor Larry Maloney. The first part of this conversation is available here.
Maloney: What is the focus of your own field robotics company, RedZone?
Whittaker: We concentrate on one principal market, robotic inspection and maintenance of water and sewer networks, which is an incredibly large market opportunity. Our robot systems combine closed circuit TV with synchronized laser and sonar technology to determine aging and detect problems, such as pipe corrosion. Our line of robots range from small inspection models to large robots that can do repair work.
Beyond this vertically integrated market, there’s the whole area of nuclear waste disposal. Early robots developed at RedZone did inspection and cleanup work after the 1979 nuclear reactor accident at Three Mile Island in Pennsylvania. In the years ahead, there will be a significant market for robots in what the Department of Energy (DOE) calls environmental management, which is essentially the cleanup of legacy nuclear facilities from the Manhattan project and the Cold War.
Maloney: On the subject of government funding, how important are initiatives, such as DARPA’s (Defense Advanced Research Projects Agency's) series of Challenge events, in fostering field robotics technology?
Whittaker: They have played a very important role, and our teams at CMU have participated in many of them, including the recent DARPA Robotics Challenge, a competition of robot systems capable of assisting humans in natural and man-made disasters. Our teams have done very well over the years, including winning the $2 million prize in the 2007 DARPA Urban Challenge, in which autonomous vehicles negotiated city and suburban streets. So DARPA has certainly provided invaluable support, especially in the early days of field robotics when researchers were developing the technologies for driverless ground and air vehicles.
But we shouldn’t stop at DARPA. The National Science Foundation, NASA, DOE, DOT (Department of Transportation), USDA (U.S. Department of Agriculture) and even the National Institutes of Health have all provided backing for robotics research. NASA’s Space Robotics Challenge, for example, is encouraging continued work on the Valkyrie, a magnificent state-of-the-art humanoid targeted for future space missions. This widespread government support tells you the degree to which robotics has become relevant across so many sectors of our economy and society.
Maloney: As you look ahead, what’s the potential for even greater growth in field robotics?
Whittaker: The applications are limitless. The reason why field robotics has been a winner is because it serves vast markets, such as agriculture, mining, automotive, defense, space and infrastructure. When an airliner goes down over the ocean, robotic systems locate the plane and bring up the black box. Ideas for new applications are cropping up all the time. I’ve been thinking, for example, of a robotic system that would install fence posts across the vast livestock ranges in the West. And in space, robots have a huge role to play in exploration.
Maloney: How long will it be before we see driverless cars on the roads?
Whittaker: It’s not when will we see them; it’s when did we see them. To a great extent, driverless cars are already here, given the steady introduction of more and more automated features. It’s an evolution.
If you rely on an active park-assist system to parallel park your car, you’re already ceding control to automation. The same goes for collision avoidance systems that will put on the brakes if you’re backing blindly into traffic. Automotive OEMs are investing enormous sums in developing automated systems, and they will strategically introduce more and more such features, based on a whole range of business factors, from costs and consumer expectations to legal and regulatory issues. The totally driverless car? 5-10 years.
Maloney: What technologies are vital to future advances in field robots?
Whittaker: Despite all the technical progress we’ve made, field robotics is still only about 5% up the ladder of evolution. Still, engineers and scientists in this field are very quick to build on proven technologies. A generation in robotics is much, much shorter than the 20-year generation in a family.
Moving forward, we need to see continued improvements in sensors, especially for applications in difficult weather conditions, such as rain and fog. There’s also a demand for specialized sensors that can provide higher resolution and accuracy. Improving the reliability of robotic systems remains a big challenge, especially in such areas as computing, electronics and software. Many field robots operate in mission and safety-critical applications. That means that in many cases robotics will need to adopt the kind of triple redundancy in software and functional components that we now see in defense and space work.
Beyond that, we need to see more innovation in robotic platforms. What does the robot look like that automatically installs fence posts? And what kind of form does a robot take to descend into the pits and caves of the Mars? Finally, we need a greater evolution of artificial intelligence as applied to field robots – the ability for these machines to think.
For More Information
Carnegie Mellon University Robotics Institute: https://www.ri.cmu.edu/
CMU Field Robotics Center: http://www.frc.ri.cmu.edu/
RedZone Robotics: http://www.redzone.com/
Astrobotic Technology: https://www.astrobotic.com/
Video on the Dante volcano explorer: https://www.youtube.com/watch?v=JJ5mx5-PHIs
DARPA Robotics Challenge: http://www.theroboticschallenge.org/overview
NASA Valkyrie humanoid robot: http://spectrum.ieee.org/automaton/robotics/humanoids/nasa-wants-help-training-valkyrie-robot-to-go-to-mars
Video on drones for commercial use: http://www.youtube.com/embed/lUCYuvQ3ccA?rel=0&showinfo=0&autoplay=1