Transcript Mining and Robotics: A Next New Thing Joins the Commodities Bull

Mining and
Robotics
A next new thing joins the
commodities bull market
Presentation to the Northwest Mining Association
December 10, 2004 by Bill Fox
Mining and Robotics Overview
• Characteristics of Robotics.
Robotic trends, phases of
automation, variations in applications, strategic nature of robotics
• First Generation Applications. Involves modification of
conventional devices such as Load-Haul Dump (LHD) vehicles and
drilling equipment.
• Second Generation Applications.
Creation of
dedicated tele-operated robots such as Placer Domes’ MiniMole or
Sandvik Tamrock ARM 1100
• Third Generation Paradigm Shift.
Autonomous
robots in an exclusively robotic environment, such as the continuous
robotic excavation and processing of ore in deliberately flooded mines
10 km underground.
Robotic Trends
• Global mobile robotics
market expected to expand
from $5 billion to over $70
billion by 2025
• “Moore’s Law,” involves a
doubling of processing power
about every 20 months. This
is a key driver
• Robot modules are steadily
coming down in price while
increasing in capabilities,
although each at different
rates. At some point this
creates economic applications.
The “Robolution” incorporates electronic infrastructure created
during prior telecommunications and automation phases. The first
phase has been accomplished, the others have lagged.
[Soure: 1992 overview by Hatch Associates for Industry Science and Technology of the
Canadian Federal Government]
What is a robot?
• Machine capable of behavioral
sequences, sensing environment,
self-propulsion (for mobile vs.
fixed robotics), and artificial
intelligence
• Modules: Frame (skeleton) ,
Power System, Actuators (muscle),
Drive Train (feet), Controllers
(brain) , Sensors, End Effectors
(working hands), Communication,
Outer Shell (skin)
[Pictures: Tetra Vaal, The Embassy Visual
Effects, www.theembassyvfx.com; iRobot’s
Ariel robot crab, www.irobot.com]
Variations in applications
• “Robotization” is invading all areas of
the structure of production in all
industries. For example, Newmont’s
automated assay labs at Carlin, NV,
Australia’s Universal Dragline innovation
• “Robotization” builds on “sensor
suites” embedded in equipment. A
typical Load Haul Dump (LHD) vehicle may
be outfitted with 150 sensors
• “Robotization” ultimately means the
reconceptualization of mining jobs
and operations
• Japanese are spending over $100
million a year on robotic R&D
[Upper left: Intuit Surgical’s teleoperated Da Vinci
System. Lower left, a Load-Haul-Dump vehicle that uses
a light rope for autonomous navigation. Picture courtesy
of MD Robotics.]
One operating system can serve many
industries. Is there a macro-strategic role
for robotics in general, and an overlap with
robotic mining applications in particular?
Another litmus test question: “What if I
wake up one morning and learn that a
hostile country has one million robots
with human-level intelligence, which can
design and manufacture even more
intelligent robots at an increasing rate and we have nothing to compare with
this?”
First Generation Robotics in Mining:
Adapting conventional devices and work processes
[photo source: Automated Mining Systems/MD Robotics, automatedmining.ca]
Teleoperation
• Objective: run existing
equipment longer, faster,
and cheaper
• Applied to drilling
equipment, Load-HaulDump (LHD) vehicles,
and trucks
• Examples: Inco’s Stobie
mine in Canada, LKAB’s
Kiruna mine in Sweden,
and WMC Resources’
mine in Australia
[Diagram courtesy of MD Robotics.
Please visit automatedmining.ca]
Teleoperation
• It is much more than a
person working pedals and
a joystick
• Objective is to shape
specific behaviors with
artificial intelligence,
increase autonomy
• Each teleoperator controls
an increasing number of
machines
[Pictures courtesy of MD Robotics.
Please visit automatedmining.ca]
Advantages
• Teleoperation reduces machine down time caused
by worker transit time into and out of the mine.
More machine time = more throughput
• Reduces wear and tear on machines when artificial
intelligence programs shape optimal behaviors
• Increases worker safety
• Only .1 second delay over 600 mile distance
Disadvantages
• Does not help much if it does not free up a
bottleneck. Many bottlenecks are related to such
factors as frequent blasting schedules or
maintenance breakdowns that may be out of the
reach of current robotic technology.
• Up front capital costs
• Does not solve organizational problems if a
company suffers from poor leadership and general
inefficiency to begin with.
Case Study: LHD Automation
• Installation cost: $400,000, capital cost $2,000,000, total
cost of $2,400,000
• Extend LHD usage from 5 hrs to 7 hrs in an 8 hour shift,
one hr left for maintenance. LHD runs 3 shifts a day, 21
out of 24 hours.
• Use of LHD increases to about 2,160 hours a year, so we
must hire an extra maintenance laborer at $46,000 a year
plus $50,000 extra in parts
• Achieve $3,007,568 extra revenue from extra 2,160 hours
at some certain commodity price
• Payback 9.9 months, IRR 68% on five year pro forma
• Question: What would the alternative have been to have
simply hired extra labor to overlap four shifts a day?
Typical trade-off: to pay for more
human labor vs. pay for robots
• Total cost of ownership
(“x” axis)
• Vs Size of facility (y
axis)
• Robots initially more
expensive, but lower
overall cost over higher
volume or longer
period of use
[Source: INtelliBot. Please
visit intellibotrobotics.com for
product information]
Technological
constraints
• LHD can tram and unload
autonomously, but needs
guidance when loading from a
muck pile to avoid hitting
walls
• Limited learning ability to
memorize turning points in
corridor intersections and total
distance of trip. Still heavily
reliant on dead reckoning and
constrained by limited shape
recognition.
[Diagram courtesy of MD Robotics.
Please visit automatedmining.ca]
3-D visual mapping
MD Robotics’ Instant
Mine Modeler (iMM) uses
a pair of cameras to create
a 3-D virtual model
(below right) inside a
computer based on
ambient light. The system
can measure distances
from the camera to
hundreds of thousands of
points, and also creates 3D mine maps.
[Pictures courtesy of MD Robotics.
Please visit automatedmining.ca]
Laser navigation and
mapping
• Lasers can determine the
dimensions of drill holes
and other dark cavities
• More expensive than
cameras for visual
mapping
• With a robot, you can
have both lasers and visual
systems
[Picture courtesy of MD Robotics.
Please visit automatedmining.ca]
The human brain makes manipulation and navigation seem
vastly easier than arithmetic, when in fact it requires vastly
more processing power
[diagram: courtesy Dr. Hans Moravec from his book Robot: Mere Machines to Transcendent
Mind]
Teaching robots manipulation is still very hard
• Most robots currently
have the manipulation
and navigation
intelligence level of a
two month old human
infant
• Can recognize faces, but
not age progression
• Very limited ability to
pick up simple objects
[MIT research robot Cog
playing with a slinky toy,
copyright Peter Menzel, please
see his book Robo Sapiens:
Evolution of a New Species/MIT
Press]
Second Generation Mining Robots:
Designed as robots for hostile environments
Test run of Placer Dome’s MiniMole (below).
[source: please visit www.placerdome.com]
Placer Dome’s official
disclosure:
“One of the most exciting projects under
way is the development of a novel machine
that will allow the `surgical’ mining of
narrow vein and reef-type deposits. This
mechanical mining method is designed to
improve safety by removing miners from
the active rock face in underground mines.
It also aims to drastically reduce the amount
of development and waste dilution that is
experienced with conventional methods.”
Modular Dimension
• The “end effector” (the part of the robot that
does the work) can “pull” on the other
modules to make the concept workable.
• MiniMole probably exploits new
Oscillating Disc Cutter technology for
drilling rock. ODC is mounted on Sandvik
Tamrock’s ARM 1100 teleoperated robot.
Modification of Work Concepts
• “Turn it on and walk away”
slogan of the Roomba robot
vacuum cleaner. So what if
MiniMole takes longer to
drill if you have lots of them
working continuously, and it
is not on “human” time?
• Placer Dome can now use
this concept to put more
emphasis on growing its
reserves by the drill bit
[photo: iRobot’s Roomba. Please visit
irobot.com for product information]
Other second generation
characteristics
• Robots are still teleoperated
• Although each robot might perform finite tasks,
they can act like an ant colony or a mobile
assembly line in the field to achieve continuous
process flow from excavation to final product.
• Robots can carry myriad sensors and be the tip of
the iceberg of a database management and analysis
system
Third Generation Paradigm Shift
[scene: Captain Nemo shows his guests strange new things aboard the
Nautilus]
Why lasers under water?
• Dr. Greg Baiden, who holds the Canadian Research Chair
in Mine Automation and Robotics, is exploring using
lasers rather than cords to teleoperate submersible robots
so that they do not tangle in trailing control lines.
• In open air applications, lasers might supply more
bandwidth to open pit robots constrained by governmentmandated frequency limits
• Lasers might make it feasible for submersible mining
robots to work in deliberately flooded mines deeper than 5
km. Water pressure can offset tremendous earth pressures
that cause exploding rocks and also overcome increasing
heat problems at greater depths.
Making a friend out of an
enemy
• Water has historically been an enemy to pump out.
Submersible robots make it a friend.
• Robots can travel in 3-D to ferry ore or can be
connected to long hoses to pump it out in a slurry.
• Submersion requires a completely robotic
environment. Opportunities exist under the ocean,
in already flooded mines, and in underground
mines so deep it makes sense to flood them for
submersible operations.
Possible Objection:
Lasers are line-of-sight.
Also, steady electrical power may be needed for
long periods that require a cord connection.
• Possible solution idea: Create a “mother ship”
connected to long electric cords, teleoperation cable, and
other possible power sources. Use it as a docking
station for smaller worker bots to recharge. The robots
can relay laser signals to each other
• Another possible solution idea for the cord problem:
Turn the cords themselves into robots as snakebots.
Each segment carries a computer chip, allowing it to
recombine with other segments in new ways.
Strategic Issues
• The mining industry offers
developmental challenges
relevant to general robotics. The
industry will likely be flush with
cash from rising commodity
prices to support R&D that both
helps itself and supports broader
strategic goals.
• Regarding autonomous vehicles,
a few years ago Congress
mandated the U.S. Army to have
as its goal one robot for every
third Army vehicle by the year
2015
[Picture: artist concept at DARPA site
darpa.mil/grandchallenge04/index.htm]
Social Issues: Robots and
Humans Should Be Friends
• Automation expands the division of
labor, increases the number of jobs
and wealth over the long term
• On balance, the alternatives are
worse
• Technology is only a form of leverage,
and while it can make social problems
more solvable, it is up to people
themselves to solve their own
problems
• The character of automation reflects
the character of the underlying
society
[Forbidden Planet’s Altaira Morbius and
Robby. Copyright 2004 Anne Francis.
Please visit www.annefrancis.net ]
“The U.S. needs to change its culture so scientists and
engineers, not athletes, are heroes. [If it doesn’t] this
country will continue to get what it celebrates.” Dean
Kamen founder of F.I.R.S.T.
• F.I.R.S.T and ROBOlympics competitions have had great
success stimulating grass roots interest in robotics
• Mining companies should provide incentives for their
employees to embrace the “robolution” and initiate their own
pet 2nd generation projects like Placer Dome’s MiniMole. They
should even consider reaching for the 3rd generation.
• Some day it would be a good thing to have a mining industry
“Robot Hall of Fame” to honor robot developers and their
creations.
For more information
• To learn more, see my article
on the web “Mining and
Robotics” and also my
overview article on the mobile
robotics industry.
• Please refer to my web site
www.amfir.com, and click on
“Bill Fox Archive” to see
other articles and works in
progress.
• Also, please visit web sites for
MD Robotics, Penguin ASI,
and SIAMtec.
[Montage: The many activities of
Automated MiningSystemts/MD
Robotics, www.automatedmining.ca]
Special thanks for assistance in
researching this project
•
•
•
•
•
•
•
•
Placer Dome Technical Services
Newmont Mining Technical Services
MD Robotics
Penguin ASI
SIAMtec
Eric Jackson, Cellula Robotics,
Vancouver, B.C.
Dr. John P.H. Steel Colorado School
of Mines
Prof. E. M. Nebot, Australian Centre
for Field Robotics, U. of Sidney
QRIO image courtesy of Sony Corporation