William L. “Red” Whittaker is director of the Field Robotics Center and founder of the National Robotics Engineering Center at Carnegie Mellon University in Pittsburgh. His expertise includes developing robots for hazardous duty and for performing 3-D mapping and remote sensing in environments such as coal mines and volcanoes.
Soon after a near-fatal mine disaster at Quecreek, PA, in mid-2002 (in which nine miners using an outdated map mistakenly broke through a rock wall, flooding their tunnel), interest grew in his center’s subterranean robotics and its mapping capabilities. A prototype machine they’d built, an autonomous, four-wheeled robot with heavy-duty tires, called Groundhog, was sent into an abandoned coal mine near Pittsburgh in May 2003, and was able to create accurate three-dimensional maps of its surroundings. It proved the ability of a robot to map the rooms, pillars, and corridors left by generations of mining.
In the aftermath of the January 2 disaster at the Sago Mine in West Virginia, which cost the lives of 12 miners, Technology Review asked Whittaker to discuss the possible role of robots in aiding and rescuing miners. Whittaker did not discuss the specifics of the Sago disaster, but instead spoke about the potential of underground rescue robotics.
Technology Review: What could robots do in mine rescue situations?
Red Whittaker: First, it’s important to realize that, although the technologies exist, robots are not yet certified or deployed as standard tools in mine rescue today.
Rescue robots in the future will certainly enter mines – under the unknown conditions of dust and gas and inundation and roof fall – and will be crucial for exploring and characterizing conditions and reporting back to command centers. They can carry gas sensors that characterize the atmosphere of a mine. Typically, they would deploy two of those sensors on each machine to make sure there is no mistake in the instruments.
Once robots have the capability to get in and get around, they could also provide communications and visual and map sensing, deliver objects to aid trapped people, like oxygen tanks, and detect vital life signs.
TR: What is the state of the technology?
RW: The state-of-the-art technology for automated navigation in a mine -– a robot sensing and planning and driving and getting from point A to B, knowing where it is, and staying out of trouble, and getting back to an egress point – is pretty well understood. Further, robots can carry sensors that would alert people to the presence of methane gas or poisonous air. Also, in most mine accident response scenarios, there is a prior map – that is tremendous information for a robot and a rescue crew.
TR: So robot locomotion in a mine is practicable?
RW: In a mine there are corridors and intersections and walls and floors and a roof – for a robot’s navigation and reasoning that’s a lot of information. But it’s still a lot different than sending a robot into rush-hour traffic, for instance, to head two miles across town. A mine is a relatively simple world for a robot because it is uncluttered by many unanticipated items. In an office building there is far more complexity and clutter – desks, water coolers, signs, and people. So a mine is an amenable environment for a robotic device designed for simple navigation.
There is no fundamental barrier to good locomotion or moving through mine conditions or getting command and control via that robot or appending sense detectors or illumination devices or scanners. So useful rescue response robots could be specialized and deployed in the near term – there’s no leap of physics or big missing piece of technology for machines that could move quickly and effectively in mines.
TR: What will it take to deploy rescue robots?
RW: All the things that help any technology develop: market forces, political pressure, humanitarian impulses, and teams committed to the challenge. Five years ago the technologies would not have been competent for or capable of mine mapping; but after Quecreek there was a motivation for progress. The great strides in mine mapping since then are powered by resources that were a response to what was viewed as a shortfall – a need. It didn’t just come to pass that old pencil-drawn maps of mines were digitized, it became a necessity. There was a political, business, and humanitarian impulse at work that led to the creation and deployment of robotics to map mines, to promote the safety and health of the people in the mines.
After Sago, the charter might be for capabilities to enter mines robotically in accident conditions and traverse and obtain information and get to trapped miners and deliver what’s needed to them.
TR: Do you think that will happen?
RW: Twenty years ago this would all have been science fiction; but now it is a matter of integrating all the existing systems. Robots are now a tool of the trade for bomb squads. Ten years ago we wouldn’t have been talking about that as a reality. So the issue is no longer whether the technology will work, but of culture, policy, economics, and initiative. Every technology has to earn its keep – it’s not an entitlement. I believe that mine rescue response is one application where robots would inevitably make good sense.
TR: You grew up in coal country. What was your reaction when you learned of the Sago disaster?
RW: Any time I hear about a mine accident my first reaction is human and my thoughts are related to hope for the wellbeing of the people. My work is heavily motivated by my own background as a Pennsylvania native who grew up near the Quecreek mine. But mining accidents and incidents are not unique to any corner of the world – entrapments and natural disasters are world issues, not backyard issues.
The big new idea for making self-driving cars that can go anywhere
The mainstream approach to driverless cars is slow and difficult. These startups think going all-in on AI will get there faster.
Inside Charm Industrial’s big bet on corn stalks for carbon removal
The startup used plant matter and bio-oil to sequester thousands of tons of carbon. The question now is how reliable, scalable, and economical this approach will prove.
The hype around DeepMind’s new AI model misses what’s actually cool about it
Some worry that the chatter about these tools is doing the whole field a disservice.
The dark secret behind those cute AI-generated animal images
Google Brain has revealed its own image-making AI, called Imagen. But don't expect to see anything that isn't wholesome.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.