Robot dog fills mining roles too dangerous, boring for humans Metal Tech News – April 28, 2021
Boston Dynamics' robot dog Spot has the agility, perseverance, and skills to be a miner's best friend by taking over the risky and mundane mining tasks.
Spot is already carrying out patrolling duties at mines around the globe and the Colorado School of Mines recently unboxed its own robot dog.
"We have never had a robot that is so agile, that is so well designed, that can do a lot of things just out of the box," said Hau Zhang, an associate professor in the Department of Computer Science at the Colorado School of Mines.
As founder of the Human-Centered Robotics Laboratory at the prestigious mining school, Zhang knows a thing or two about robots and their potential applications to the mining sector.
The same can be said for Boston Dynamics, which was founded on the idea of coming up with ways that robots could navigate a world that is unstructured, unlevel, or dynamic.
"A world where you can't expect flat, consistent surfaces that wheeled and tracked robots typically navigate easily," Michael Perry, vice president of business development at Boston Dynamics, told Metal Tech News.
"Having legs enables you to have what we call 'human-like' mobility – if you can walk on it there is a good chance the robot can handle it as well," he added.
This mobility is what makes Spot a great companion around a mine, whether that is climbing over a rockpile to check the safety of a mine that extends nearly two miles underground or climbing stairs while doing a preventative maintenance check of the mill on the surface.
"We have a fleet of tracked and traditional wheeled robots that we use underground but there is nothing that compares to the way Spot steps or moves around objects," said Iain McKillip, technical services manager at Glencore's Kidd Creek copper-zinc-silver mine in Ontario, one of the first mining operations to utilize Spot's capabilities.
For Zhang, however, there is another more fundamental reason other mines should consider using Spot.
"First of all, it is cool!" he said.
Secondly, the Colorado School of Mines professor says robots such as Spot have the ability to inspect areas that are too dangerous and sometimes even impossible for humans.
The Spot robot at Kidd Creek is earning its hazard pay.
Extending to depths of 9,600 feet, or about 1.8 miles, below sea level, Kidd is the deepest base metal mine on Earth and boasts the longest surface-to-bottom underground ramp.
Spot does not protect its human companions by simply descending to great depths. Instead, it is what the robot is tasked to do after it descends into the mine's depths that keep people out of harm's way – inspecting underground development faces after a blast.
While human inspectors would need to wait until the blasting gasses have been vented out of the area, Spot does not have those concerns and can send back images and data much earlier.
While it is there, Spot can also check to ensure all the blast charges have detonated before crews arrive to dig deeper into the Kidd Mine.
This means that mining crews can more quickly and safely get to their jobs of digging out the blasted rock once the air has been vented.
"The story there is crystal clear – rather than risk a person to do a very simple task, send a robot instead," Perry said.
The Spot robot at Kidd also carries a payload of gas, seismic, and laser sensors to help ensure the integrity and safety of the mine while carrying out its underground patrols.
Aerial drones could carry out similar tasks, but Spot offers two key advantages – payload and range.
Spot has the ability to carry and power roughly 30 pounds of sensors and scanners and pack those around for 90 minutes, dependent on the power draw of the accessories it is carrying. If Spot's sentry duties require longer durations, the robot can simply get recharged at self-docking stations set up at centralized locations or intervals along the path.
Perry says that while Spot usually gets hazard duty as its first job in a mine, the robot is often assigned less dangerous tasks as crews and management discover new ways to utilize its data collection abilities.
One such job that Spot excels at is routine inspections that tend to be boring and repetitive for humans.
Where a human may begin to relax his diligence after several hundred times of inspecting the same valve that has not been a cause of any problem in the five years since it was installed, the robot will apply the same observation protocols on the thousandth inspection as it was programmed to do on the first.
And, with a robotic arm recently developed by the Boston Dynamics team, Spot can even adjust valves, flip switches, or make other adjustments based on valve readings or human inputs. The arm also comes in handy for moving objects and gathering data that is otherwise out of reach of the robot dog.
In addition to not getting bored, robots are really good at delivering non-biased and consistent datasets.
One such task being considered for Spot robots is tailings dam inspections. What makes Spot so good at a job like this is it would make the same photographic, lidar, seismic, and other observations using the exact same setting from a precise location on an ongoing basis. For mine managers, this means that anomalies that may offer early signs of problems could be more readily detected through the comparison of accurate and consistent data collected over days, weeks, months, and years.
"Why not pass that task to the robots so that people can focus on the decision making based on the robots acquired?" Perry queried.
Another non-hazardous mining job that has emerged for Spot is a preventive maintenance inspector of stationary equipment. While industrial internet of things sensors are being developed for this task, a large mill complex would need thousands to collect the same data as a single Spot unit.
Equipped with microphones to listen for grinding and other acoustic anomalies that are tell-tale signs of worn parts, thermal imagers that can tell if a part is getting too hot, and other sensors, Spot can let mine management know when parts are getting worn. This data assists in scheduling maintenance and repairs at a time that works with mine scheduling.
One of the first preventive maintenance jobs that emerged for Spot was inspecting conveyor belts for defective rollers and other problems that could put the materials transport system out of commission.
In essence, Spot has the capacity to be the golden retriever of the mine site, only in this role fetching data.
"We think that we can make data acquisition cheaper, more diverse, and more consistent," said Perry.
This is something the technical team at Kidd Creek has discovered.
"When it comes to engineering driven mining like we do, it is based on data – it is driven by our ability to get information on a regular basis and turn it around so that we can make good decisions based on that information," said McKillip.
"Tools like this get us right up close and personal with the things we need to see," he added.
And of course, there is Professor Zhang's initial observation – Spot is cool.