AUTONOMOUS VS MANUAL HAULAGE TRUCKS - HOW MINE SIMULATION

CONTRIBUTES TO FUTURE HAULAGE SYSTEM DEVELOPMENTS

Juliana Parreira, John Meech

Norman B. Keevil Institute of Mining Engineering,
The University of British Columbia, Vancouver, BC, Canada, V6T1Z4
Corresponding author: (jpareng@hotmail.com)

Abstract
The past decade shows record increases in global demand for minerals and metals with an
accompanying significant decline in skilled personnel. Mining companies today are examining
new technologies to maximize production, reduce costs, and create safer work conditions to deal
with this shortfall. Automation of an open-pit mine haulage system is receiving attention as a
beneficial option since it provides more consistent and efficient operation of mining equipment, it
removes workers from potential danger, it reduces fuel consumption significantly reducing
greenhouse gas (GHG) emissions, and it can help optimize vehicle repairs and equipment
replacement because of more-predictable and better-controlled maintenance. This paper describes
a simulation model under development using the EXTENDSIM software package to examine
scale-up issues and constraints of autonomous vehicles in comparison to a manual system. The
software can predict and manage Key Performance Indicators (KPIs), such as productivity,
safety, cost, equipment failures, fuel consumption, and tire wear under different road and load
conditions. By breaking-down the various systems of an autonomous haulage truck and
simulating the overall haulage system using a Monte Carlo approach, the model becomes a
flexible and powerful tool to use within a mining organization. Direct comparison to a manuallyrun system can be done to show the cost and operational benefits of automation. Project
managers can use this tool to guide decision-making about possible application of autonomous
haulage trucks at a specific mine site.
Keywords: Discrete Events, KPIs, mining industry, management, autonomous haulage trucks.

Growth of Automation
Todays organizations face many challenges: demanding customers, global competition, and
currency fluctuations. As a result, organizations are looking for automated processes to improve
efficiency. According to a study conducted by the ARC Advisory Group, (a leading organization

CIM Meeting, Vancouver-BC, Canada 2010

How Simulation Contributes to Future Haulage Systems

in manufacturing, logistics, and supply chain solutions) process automation systems are expected
to grow globally by 9.6%, at a compounded annual growth rate with an contribution of over $47
billion by 2011 to the world economic markets. Asia is leading in large automation projects, with
China driving the global demand [1].
Globalization is the main factor that compels growth of process automation systems, because as
opportunities and partnerships expand there is increased complexity around the management and
distribution of diverse products and markets. Competition has heightened and successful
companies require improved efficiency, product-quality monitoring, product and equipment
availability, employee safety, flexibility, and delivery performance. To this end, it is important to
incorporate automation into the organizational processes of an enterprise since one of the main
characteristics of an automation system is rapid response to environment changes.

Definition of Automation
The Oxford English Dictionary (2005) defines automation as the use of machines to do work that
was previously done by people: so automation is often interpreted as the loss of jobs [2].
Parasuraman et al. defines automation as the full or partial replacement of a function previously
carried out by a human operator [3]. These two definitions suggest that the objective of
automation is to control the behaviour of dynamic systems and emulate physical and intellectual
human capacity. There are several degrees of automation and these can be simplified into the
following categories: direct control, supervisory, automatic and autonomous systems, where there
is less human intervention during the operation with each successive degree of automation [4].
Mining automation has employed direct tele-operation; for example, workers control the mining
process using a computer in a control room; remote operation; for example, drilling is performed
by workers using joysticks from a distance; and autonomous; for example, open pit/ underground
autonomous haulage trucks with no human intervention.
Although the Oxford Dictionary definition claims the main focus is worker replacement, as
Bibby et al have pointed out even highly automated systems during the 1975 energy crisis, such
as electric power networks, needed human beings for supervision, adjustment, and maintenance.
So, one can draw the apparently paradoxical conclusion that automated systems are man-machine
systems for which both technical and human factors are extremely important [5]. Automation
2 CIM Meeting, Vancouver-BC, Canada 2010

does not always result in people replacement; on the contrary, it can elevate human capacity into
higher level decision-making. Whatever the case, workers need a different set of skills to handle
the specialized tasks related to automation and the new technical challenges.

Autonomous Haulage Trucks (AHT)

In mining, automation is playing an increasingly important role due to the scarcity of high
demand metals; challenging locations and harsh environments becoming the norm. As a result,
automation is being used increasingly throughout the industry from underground to open pit
mining in order to increase process efficiency, safety, and production.
Due to advances in automated tracking systems, control equipment, telemetry and robotics, there
are major improvements expected in the precision and safety of mine machinery. As a result of
these advances, driverless haulage trucks are now being developed for open pit mines. Wireless
communication, object-avoidance sensors, on-board computers, GPS systems and artificial
intelligence approaches enable haulage trucks to drive themselves, or to be driven by an operator
at a control panel well-away from any danger [6].
Komatsu and Caterpillar are the two companies responsible for autonomous haulage truck
development. Komatsu is already using this technology in Rio Tintos Pilbara mine, an iron ore
mine located in Australia. In December 2008, Komatsus FrontRunner Autonomous Haulage
System started trials at this mine with driverless trains and trucks controlled from Perth, 1,300
kilometres away. All truck navigation at the mine is remotely controlled [7]. At the Codelco/Rio
Tinto Gabriela Mystral mine located in Chile, a second Komatsu project has been installed [8].
The Caterpillar autonomous haulage truck project is a joint-venture project with BHP-Billiton. As
part of the alignment, Cat and BHP Billiton are launching joint autonomous mining haulage
development programs. The companies development program includes enhancing the existing
mining trucks by integrating them with robust autonomous sub-systems many of which are
Caterpillar technologies already proven in the marketplace. The companies plan to have
autonomous trucks at selected mine sites by 2010 [10].
Autonomous haulage trucks contribute greatly to reducing losses associated with human elements
such as individual performance, personal breaks, and absenteeism [9]. While driverless haulage

How Simulation Contributes to Future Haulage Systems

trucks are not immune to breakdowns, increased consistency and scheduled maintenance will
increase the lifetime of machine components, leading to longer periods between maintenance,
and therefore reduced costs associated with maintenance. Lost production can be minimized or
eliminated as unpredicted breakdown frequency declines [10]. Lost time is not always accounted
for as a specific cost since mines often contract equipment in order to meet production targets.
Fuel consumption is reduced when a truck is driven in a stable, consistent manner. Drivers
perform well at the start of a shift but as fatigue sets in towards the end of a shift, their
performance varies considerably. By reducing fuel use, greenhouse gas emissions, and operating
costs, autonomous haulage trucks directly contribute to the principles of sustainability.
Statistics reveal that using autonomous haulage trucks in an open pit improves safety,
maintenance and equipment life, optimizes fuel consumption, and provides streamlined
operations with increasingly accurate production systems. But even with these advances, mining
companies must work hard to connect this new technology to other organizational processes. The
decision to implement an autonomous process in a mine must consider all possible impacts, not
only operational improvements.

Implementation Strategies
The success of an automation project depends on good implementation strategies, commitment
and vision of the organization, and sponsorship. It is important to identify the Major Critical
Success Factors (CSFs) related to an autonomous haulage truck project. Examples of CSFs to
describe an automation project are as follows:

Formulation of a simplified and standardized project plan prior to project design;

Assembly of personnel teams with the necessary experience and motivation to properly
implement the system;

Development of the functional structure of operations, maintenance, and management as

modular units;

Effective communication within these functional groups and documentation that clearly
spells out the plans and objectives;

Evaluation of the performance of the system and modifications done to keep it operating
efficiently.

4 CIM Meeting, Vancouver-BC, Canada 2010

The use of a Balanced Scorecard (BSC) is a tool that allows an organization to devise and
monitor CSFs enabling the successful implementation of a complex project [11]. A BSC
evaluates an organizations performance in a number of different areas with the objective to align
new business activities to the existing and new visions and strategies. It assists in improving
internal and external communications, and in generating continuous monitoring of performance.
AHS will be successfully accepted in a company and across the industry if the Balanced
Scorecard shows clearly how the CSFs link to performance measures of project phases such as
implementation, start-up, and operation. Knowing how to monitor performance measures,
especially throughout transitional and implementation stages, is important. An organization must
monitor and assess these indicators continuously since variable results will impact and/or develop
best-practices for decision-making, organizational strategy, quality-control, measurement, and
reporting improvement trends of the performance indicators (PIs). Certain indicators are key to
the bottom-line aspects of the overall project implementation and success (KPIs) while others are
essential in anticipating future problems and necessary changes.
To this end KPIs that represent the set of selected measurements must focus on critical
performance aspects. They must be analyzed and monitored on a regular basis to quickly identify
output changes that emerge as a result of any technological changes. In implementing an
Autonomous Haulage System, for example, it will be useful to know the degree (or level) to
which a robotic system can approach or exceed the productivity of a manual system as well as the
control factors that affect productivity.

Importance of Predicting KPIs

To predict the level of improvement of AHS, simulation software that analyzes the behaviour of
this new technology can be used to study possible changes in adaptability and utilization.
Komatsu uses software called Modular's Dispatch system that categorizes haul truck time into
five major categories: ready, down, delay, standby, and shift change. From these classifications,
key performance indicators such as haul truck fleet utilization may be calculated [12].
KPIs are not past indicator measuring events [11] and so, simulation software help to predict
benchmarked KPIs as well as discover new ones that might characterize changes from new

How Simulation Contributes to Future Haulage Systems

technology. This will quantify improvements and enable decision-making, i.e., the level of
improvement to justify the cost and benefits of the technology.
Technology alone is only one piece of a complex solution to improve performance. Significant
benefits will be achieved when technology, processes, and possible challenges are integrated with
the key success factors and performance measures used to guide planning and management.
Failure to properly incorporate success factors and KPIs may result in failure of the system or
faltering authorization from top management. To this end, offline simulation software that allows
prediction and comparison of KPIs according to autonomous and manual haulage systems should
be incorporated into the project.
The main purpose of simulation is to improve the technology and make sure that the autonomous
haulage trucks adapt smoothly to the organizational process, adding value. The software can
show the cost and operational benefits of automation and in cases when autonomous trucks KPI
changes are unexpected, it can be used to develop improvements in system utilization. In
addition, the tool can help to develop and analyze new KPIs for this new technology.
Simulating AHS allows the determination of overall improvement in the long term across all
KPIs. The presence of certain attributes or absence of specific constraints at a particular mine
may be necessary to ensure the overall performance improves, i.e., weather conditions,
topography, geology, etc., can be studied.

Simulation Software
The proposed software consists of two components: supervisory software and model. This paper
describes the models used for the simulation which are still under development.
By definition a model is an abstracted and simplified representation of a system at one point in
time. Models are an abstraction because they attempt to capture the realism of the system. They
are a simplification because, for efficiency, reliability, and ease of analysis, a model should
capture only the most important aspects of the real system. Dynamic modeling is the foundation
for computer modeling [13]. To this end, the word model in this paper describes a particular
dynamic behaviour of a system or a process. One of the main benefits of a model is that it begins
with a simple approximation of a process that is gradually refined as more detail is examined. As

6 CIM Meeting, Vancouver-BC, Canada 2010

refinements are added, the model more closely imitates the real-world process across a wide
variety of variable changes.
Discrete event modeling is used in our approach, i.e., the system changes state as events occur
and only when an event occurs; the passing of time has no direct effect on the model activity or
performance [13]. The simulation software consists of two main models: one using only manual
trucks and the other using autonomous trucks. By having these two models in the same virtual
environment, direct comparisons can be done and interaction studies performed. In addition, a
Monte Carlo approach allows relationships and parameters to vary randomly within a distribution
of possibilities. This provides a range of results rather than a single output for each test run.
The discrete event models have a set of constraints regarding real-world factors that influence
their performance. These parameters include, but are not limited to, climate conditions, haulage
road topography, dust, snow, mud, truck-performance, maintenance-scheduling, interaction with
non-robotic trucks and auxiliary vehicles, and aspects of the communication network and various
sensors. As well independent elements such as shovels, roads, road segments, dumps, digging
points, water trucks, cheery pickers, light vehicles, and dozers are being modelled as well.
The simulation models are being built in ExtendSim, a graphical software package that can break
a network down into unique components each having specific delays and characteristics with
respect to maintenance, speed, fuel consumption, braking, acceleration, etc. ExtendSim has
benefits such as ease of use, provision of debugging and error diagnostics, the ability to
communicate with other software tools such as spreadsheets and external execution programs.
Animations and a graphical environment for visualization of the simulated mining process are
also important features.

Manual Haulage Model

The manual haulage model includes sub-models to characterize the human truck drivers who
must work 12-hour shifts on a 2-week-on/2-week-off schedule with all trucks running 24/7. The
manual model is set to account for the performance of different drivers new and experienced
consistent and inconsistent recently trained or retrained ones who tire appreciably at the end
of the shift ones who brake too hard ones who play tunes on the accelerator consuming
excessive fuel etc.
7

How Simulation Contributes to Future Haulage Systems

Due to shift changes, truck and shovel availabilities, individual worker performance, weather
changes, mining schedules, and other factors, mine production may change. The manual model
will determine the base case against which all other test work will be compared with an initial
focus on shift production of material loaded, hauled, and dumped.
In addition to mechanical availability, the manual truck model considers two types of
maintenance: Failure and Planned. Exponential and Weibull distribution functions will be used to
introduce each type of failure on a random basis with increasing frequency as time passes if
planned maintenance is not conducted. The Weibull distribution is an excellent way to model
failures from mechanical degradation such as corrosion and fatigue, while other types of failures
such as refueling are better modeled using an exponential distribution. Maintenance costs will
also be included.
2

Figure 1: weibull probability density

Autonomous Truck model

This model is broken-down into various sub-systems or agents of an autonomous haulage truck
such as communication, brake, accelerator, steering, positioning system, radar sensors, and other
instruments and actuators. to understand the full interactions that may affect autonomous truck
behaviour. Behaviours will be studied and analysed for different external environments such as

8 CIM Meeting, Vancouver-BC, Canada 2010

road conditions, interactions with non-robotic vehicles, climate circumstances, etc. Simulating
different autonomous behaviour can generate knowledge to predict how a sub-system (agent) or
the overall system can be reconfigured to improve system efficiency when a step (or pulse)
change occurs in any system parameter.
Breaking-down the various sub-systems of an autonomous haulage truck and simulating these
using a Monte Carlo approach enables a better understanding of system sensitivities to different
sub-system variations. Figure 2 shows the possible sub-systems (agent) of an autonomous
haulage truck being considered in the AHT model. Direct comparison of a manually-run system
with the autonomous one can show the added costs and operational benefits of automating trucks.

Figure 2: Sub-systems of an autonomous haulage truck.

KPIs considered
Important KPIs from each model will be chosen in accordance with the benchmarking success of
the mine. KPIs include productivity, safety, breakdowns, maintenance and labour costs, and cycle
times. Additional KPIs that are important during the work include fuel consumption, tire wear
and investment per truck. Not all KPIs are likely to improve when automation is introduced. For
example, the AHT speed may be limited for safety or technical factors to a level below that of the
manual truck, although the variance will certainly decrease. Despite this reduced performance, it
is expected this will be balanced by significant improvement in availability, mechanical and

How Simulation Contributes to Future Haulage Systems

reduced waiting. The overall changes across all KPIs will be compared and the combined effect
evaluated. At this stage, the following estimates (or targets) of KPIs differences are expected:
Investment cost per truck

+30%

Truck haulage cycle times

- 7%

Fuel consumption

-10%

Tire wear

-12%

Mechanical Availability

+ 8%

Increased productivity

+ 5%

Maintenance costs

-14%

Increased truck life

+12%

Labour costs

- 5%

Improved Safety/Reliability to be determined

Note that the expected improvement in Labour costs is not particularly high. Despite there being
no further need for four truck drivers (two on each cross shift cycle) for each truck, additional
personnel will be needed to maintain the trucks and their sensor and control sub-systems. These
employees will be much more skilled intellectually and thus, will have increased salaries.

Software Architecture
With an autonomous haulage truck system, it is important to choose a software architecture that
can deal with each element in the system as an "Agent". Each truck, each shovel, each dump,
each road section, and all other vehicles in the mine must be considered as independent,
interacting Agents. This type of architecture allows maximum efficiency and effectiveness in
developing and running the overall software system. It guarantees that access to computer
hardware resources follows a consistent pattern that accounts for priorities and maximizes model
effectiveness and efficiency.

10 CIM Meeting, Vancouver-BC, Canada 2010

Figure 3: Interface Agent Architecture

An agent is an entity that receives requests for services from other agents (who are Clients with
respect to the servicing agent). If it is currently inactive, then it fulfills such services. For
example, a truck arrives at a loader and requests the loader to fill its bucket with ore. The loader
receives the request, processes it to establish a "contract" with the truck, and then fulfills the
"contract" by loading the truck.

Another example: a truck arrives at the start of a road segment and requests permission from the
main supervisory system to enter the road segment. The main supervisory system (acting as an
Agent to the truck) provides such permission and instructs the road segment to allow the truck to
enter. The truck enters the road segment and is "serviced".
Each instrument, final control element, motor, switch, light, etc. mounted on each truck can also
be considered agents. As such, Agents represent a hierarchy within the system in terms of actions
and interactions. This structure permits redundancy since logical inferences can be made about an
Agent even when it fails to communicate its performance to the central supervisory system.
A behaviour is a specific action performed by an agent or a collection of agents. For example:
A truck must stop for some reason: Stopping is Behaviour.
A truck must turn left for some reason: Turning Left is Behaviour.

11

How Simulation Contributes to Future Haulage Systems

A truck must accelerate for some reason: Increasing speed is Behaviour.

A truck must decelerate for some reason: Decreasing speed is Behaviour.
A truck must back-up for some reason: Backing-up is Behaviour.
A Task is a particular series of behaviours that must be performed by each element at any point in
time. For example: When its bucket is full, a truck must move along a haulage road from a
loading point to a dump point. When a truck arrives at a dump position, it must place itself to
perform dumping; when a truck arrives at a loading position, it must place itself to allow the
loader to fill its bucket. In order to accomplish each of these tasks, a wide variety of behaviours
must be conducted: some in parallel and some in sequence.
Tasks are repetitive operations carried out in sequence by an Agent or group of Agents. Some
Agents perform one, and only one, particular task over and over again as new "customers" arrive
to be serviced. Other Agents conduct different tasks dependent on their interactions with other
Agents and on the circumstances that change in the system (environmental, design, unexpected
events, interruptions, etc.).

Conclusion
The field of automation is changing at a rapid pace. While at first glance; automation appears to
devalue labor by replacing people with machines, in fact, automation leads to advancement in
intellectual capacity and higher-skilled mine operators. Automation leads to an improved quality
of life and better workspace for people, as risks and exposure to unsafe environments are
minimized. Automation increase equipment utilization and efficiency.
For an AHS project to be successful, organizations must identify and analyze levels of interest,
expectation, priorities and influence of stakeholders in the early stages, as well as develop a
management plan which incorporates quality control, risk management, communication plans,
and exit strategies. It is important to identify the Major Critical Success Factors (CSFs) related to
an autonomous haulage truck project and develop a Balanced Scorecard (BSC) as a management
framework to identify and monitor CSFs. Great effort should be given to performance measures,
as they allow warning alarms, target performance, and time trend comparisons both within the
organization and with others benchmarking quality records (Enterprise Performance Monitoring).

12 CIM Meeting, Vancouver-BC, Canada 2010

Performance Measure ensures the new technology is being evaluated with respect to relevant
indicators that measure both positive and negative contributions of the automated process.
Simulation software can help organizations predict benchmarked KPIs as well as discover new
KPIs to use in accommodating changes from new technology. This simulation tool can be used in
any open pit operation designed to receive autonomous haulage technology. Each mine is unique;
to this end, the autonomous truck behaviour will differ. By incorporating the levels of variables in
each mine, this offline tool can assist project managers to optimize projects resulting in more
assertive decision-making.
Acknowledgment
The authors wish to express their appreciation to BHP-Billiton Nickel West Division for the
funding of this research and for access to important data and information regarding autonomous
haulage systems.

References
[1] Plant Engineering. http://www.plantengineering.com/article/184542-Process_manufacturing_ automation_
DCS_market_remain_strong.php. accessed September 30, 2009
[2] Hornby, A.S. 2005. Oxford Advanced Learners Dictionary. Oxford University Press. Seventh Edition.
[3] Parasuraman, R., Sheridan, T. B., Wickens C.D. 2000. A Model for Types and Levels of Human Interaction with
Automation. IEEE transactions on systems, man, and cybernetics, Pt. a: Systems and Humans. 30(3).
[4] The Royal Academy of Engineering. 2009. Autonomous Systems: Social, Legal and Ethical Issues. Retrieved
Sept. 30, 2009 from www.raeng.org.uk/autonomoussystems
[5] Bibby, K. S., Margulies, F., Rijnsdorp, J. E. and Withers, R. M. J. 1975. Man's role in control systems. Proc. 6th
IFAC Congress, Boston.
[6] Lewis, M. W., Werner J., Sambirsky, B. 2004. Capturing unrealized capacity. CIM Bulletin, 97(1076), 57-62.
retrieved Feb. 12, 2010 from CBCA Business. Document ID: 547379681.
[7] Moore, E. 2009. Hands Off: New products offer the automation advantage. CIM MAGAZINE. 4, 56 -59.
[8] Jamasmie, C. 2009. Lonely Trucks in a Lonely Place: Autonomous Trucks Debut in Chiles Desert. Retrieved
Feb. 2, 2010 http://magazine.mining.com/ Issues/0904/KomatsuAutonomousTrucksDebutChileDesert.pdf
[9] Zoschke, L. and Jackson, M. 2000. Experiences to date with unmanned haul trucks in open pit mines. Annual
General Meeting, Canadian Institute of Mining, Metallurgy and Petroleum.
[10] http://www.miningtopnews.com/caterpillar-inc-announces-autonomous-mining-collaboration-designed-toimprove-mine-site-productivity-efficiency-and-overall-sustainability.html
[11] Bennink, C. 2008. Trucks & Transportation - Take Steps to Cut Vehicle Fuel Costs; a proactive approach can
add up to major savings. Equipment Today. Vol. 44, Issue 6.
[12] Parmenter, D. 2002. Key Performance Indicators: Developing, Implementing, and Using Winning KPIs. United
States. John Wiley & So, Inc.
[13] ExtendSim. 2007. ExtendSim User Guide. Retrieved April 24, 2009 from http://extendsim.com

13

Autonomous vs. Manual haulage trucks: how mine simulation

can contribute to future haulage system developments

Authors:
Juliana Parreira Ph.D. Student (Automation in Mining)
Email : jpareng@hotmail.com

John A. Meech Professor and Director of UBC-CERM3)

Email : jameech@dccnet.com

Agenda

1) Mining Automation Applications

2) Autonomous Haulage Trucks
3) Key Performance Indices (KPIs)
4) Simulating Haulage Systems
5) Challenges
6) Conclusion

2
2

Definition: Automation

Discipline to program &

manipulate equipment
Instrumentation, sensors,
final-control-elements,
control systems, networks,
and communication
Http://ssabh188.blogspot.com

3
Http://ssabh188.blogspot.com

Definition
Definition
Robots
Robots

Definition: Robot

"A reprogrammable,
multifunctional
manipulator
designed
to
A machine to
make our
move material,
parts,
lives easier
tools, or specialized
devices through various
programmed motions for
the performance of a
variety of tasks"
4
http:// www.physiologytoday.com

Definition
Integrated
Systems
Integrated Systems

Integrated Systems which emulate and

and this includes

enhance the human ability to:

Perceive

PERCEPTION
includes

Reason

REASONING
includes

Act

ACTION
includes

the Human-Machine
Sensor systems
and recognition

Artificial Intelligence
Knowledge Systems

Control Systems
Actuators

Interfaces required to
operate these systems

5
Dr. Greg Baiden Mining Conference Chile 2008

Automatic Operation
Safeguarded
and
Autonomy
DirectMonitored
Tele-operation
Tele-autonomy
Manual
Operation
Tele-operation

Robot Initiative
Initiative
Human
Controlled

Mount Keith Mine

INCO

Autonomous Haulage Trucks

Tinto from potential dangers

Komatsu/Rio
Removal of workers
Reduction of Green-House Gasses (GHGs)
CAT/BHP-Billiton (Australia)
Reduced Maintenance Costs
Improved Productivity

7
7

Autonomous Haulage System Elements

8 9

Autonomous Truck Sub-systems

Breaker Controller
Communication Equipment

Acceleration Controller

Radar Sensor

Vehicle Speed Controller

Front Wheel Steering
Angle Sensor

Obstacle Detector

Key Performance Indicators (KPIs)

KPIs give before & after pictures
KPIs assist with decision-making
Goals are easier to measure/monitor

Long-Term Commitment
10

Implementation
Manual

Robotic

te
d
a
e
r
e
c
g
g
a
l
e
ISnettap
R

KPI
(core)

KPI
(core)

?
0

% AHT

100
11

Offline Simulation

Predict and compare KPIs

Identify and analyze new KPIs
Show system utilization improvement
even
if certain KPIs deteriorate
Past
Manual System
Existing KPIs

Future
Autonomous System
New KPIs

12

Simulating Mine Haulage Systems

(state-of-the art)
Truck Fleet Size
Size of Trucks and Shovels
Shovel/Trucks Scheduling Schemes
Planned Maintenance Schemes
Routes/Digging Points

13

Simulation of AHT
(sub-systems)

Control
Systems

Sensors

Telecom

Steering

Centralized System
Localization

Brakes

Intervehicle Location
Navigation

Fuel System

Tire Wear

Acceleration
Human Presence
Obstacle
Avoidance
Design of an
Autonomous Haulage System
14

Questions to be Answered
1) Are robotic trucks more efficient?
2) What is the level of improvement of an AHT?
3) Are todays benchmarked KPIs efficient ways
to measure performance of an AHT?
4) Is there a need to create new KPIs?
5) Are all mines ready for AHT?
6) If not, which processes need to change?
15

Modelling Approach
Shovel/ Manual Truck model
Shovel/Autonomous Truck model

16
16

Simulation of AHT
Parameters
Shovel Capacity
Waiting Times
Road segments
Weather Conditions
Topography
Dumping/Loading Times

Mount Keith Mine - 2009

17

Simulation of AHT
Scheduled Maintenance
Refuelling
Tire replacement
Shift Changes
Major/Minor Breaks
Mount Keith Mine - 2009

18

Simulation of AHT
Monte Carlo Simulation
Focus on Production

Vale INCO - Brazil

19

Manual and AHT Model Output

Both models will output Benchmarking KPIs:
Productivity
Safety
Breakdowns
Cycle times
Maintenance and labour costs

Komatsu

20

AHT Model Output

For AHT model:
New KPIs
Fuel Consumption
Tire Wear
Investment per Truck
Reduced GHGs

21

in the Research
Key Sub-Questions
Performance Indicators
- Targets

AHS

AHS

AHS

AHS

AHS

+ 30%

+8%
-7%
Manual

Manual

Investment cost Truck haulage

per truck
speeds

-10%

-12%

Manual

Manual

Manual

Fuel
consumption

Mechanical
Availability

Tire Wear

22

in the Research
Key Sub-Questions
Performance Indicators
- Targets

AHS

AHS

AHS

AHS

12%
+ 5%
-5%
-14%
Manual

Increased
Productivity

Manual

Manual

Maintenance
costs

Increased
Truck life

Manual

Labour
costs

23

Implementation Challenges
Challenges :
Management decisions and strategies
Workplace changes
Cultural barriers and union beliefs
Possible Solutions :
Longterm vision and commitment
Provide training to avoid sabotage
Regular and consistent communication
Avoid Labour Replacement as an objective
24
2424

Conclusions
AHT improves safety,
Newproductivity
Commitment
and maintenance
AHT
cannotofbeAHT
seen
employee
The future
in as
mining
looks replacement
bright implementation
should be considered a priority to remain competitive and
Simulating
KPIsand
allows
improvement
ensure a safe
effective
operationmonitoring of AHT
Simulation bias must be avoided
Simulation can be used in real time to evaluate changes

25