FEATURED ARTICLES Advanced Railway Systems through Digital Technology

Heavy Haul Freight Transportation

System: AutoHaul
Autonomous Heavy Haul Freight Train Achieved in Australia

There are many iron ore rail lines in the Pilbara region, located in North-West Australia.
Global mining company Rio Tinto Limited operates a fleet of heavy haul iron ore trains
24 hours a day from its 16 mines to four port terminals overlooking the Indian Ocean. To
increase their operational capacity and reduce transportation time, Rio Tinto realized
that driverless (GoA4) operation of its trains was the way to achieve this. The company
established a framework agreement with Hitachi Rail STS S.p.A. This project was named
AutoHaul, and two companies worked closely on its development over several years.
Since completing the first loaded run in July 2018, these trains have now safely travelled
more than 11 million km autonomously. The network is the world’s first driverless heavy
haul long distance train operation.

Mazahir Yusuf
Anthony MacDonald, Ph.D.
Roslyn Stuart
Hiroko Miyazaki

Tinto’s Operations Center in Perth more than 1,500 km

away (see Figure 1 and Figure 2). The operation of this
1. Introduction autonomous train is achieved by the heavy haul freight
transportation system, AutoHaul*1, developed through co-
Rio Tinto Limited, a leading global mining group, operates creation between Rio Tinto and Hitachi Rail STS S.p.A.
an autonomous fleet of 221 heavy haul locomotives along (formerly Ansaldo STS S.p.A.). This article presents the
its 1,700 km line 24 hours a day extracting iron ore from development history and features of AutoHaul.
mines in the South and transporting it across the remote
Pilbara region to ports 300 km away overlooking the Indian
Ocean. A typical train formation includes a triple-headed 2. Development of AutoHaul
diesel electric locomotive and 240 wagons loaded with 106 t
each car. Typical train length is 2.5 km and total weight is The demand for iron ore has increased year by year and to
28,000 t. keep up with the demand Rio Tinto achieved driverless
Maximizing transportation efficiency is the first prior- autonomous heavy vehicle (AHV) operation and planned
ity of train operations and trains can depart after loading to realize driverless (grades of automation 4: GoA4) opera-
has completed without a pre-scheduled train operation tion of iron ore trains. This project, named AutoHaul, is
diagram. Complete train operation is monitored remotely aimed at unlocking significant safety and productivity
by a team of train controllers and maintainers from Rio *1 AutoHaul is a registered trademark of Rio Tinto Limited.

76.
Figure 1 — Heavy Haul Iron Ore Train
Typical train length is 2.5 km consisting of diesel locomotives and 240 ore cars with a total weight of 28,000 t.

Figure 2 — AutoHaul Geographical Map

The trains fitted with AutoHaul run in the Pilbara
region of Western Australia and are controlled
from the Operations Control Centre in Perth,
which is 1,500 km away.

ROSELLA
TO
COAST
Western Australia

Pilbara Region

EAST PILBARA
Perth

WEST
0 50 100 PILBARA
km

gains for Rio Tinto, as well as optimizing the company’s in stages and AutoHaul full deployment was achieved in
iron ore mining system by providing more flexibility and December 2018. Rio Tinto currently runs 98 percent of
reducing bottlenecks. AutoHaul, the heavy haul freight their AutoHaul equipped locomotive fleet in autonomous
transportation system, developed in collaborative creation mode covering in excess of 11 million km on their network.
(co-creation) between Rio Tinto and Hitachi Rail STS
S.p.A., was granted accreditation by Australia’s Office of the
National Rail Safety Regulator on 16 May 2018. Following 3. Structure of AutoHaul
this, an autonomous train, consisting of three locomotives
and carrying around 28,000 t of iron ore, travelled over AutoHaul is based on automatic train operation (ATO) and
280 km from Rio Tinto’s mining operations in Tom Price European Train Control System (ETCS) level 2. The sys-
to the port of Cape Lambert on 10 July 2018. Since this tem can be broken down into four main segments: commu-
initial run, driverless (GoA4) operation was ramped up nications infrastructure segment, operations center segment,

Hitachi Review Vol. 69, No. 6 788–789 77.

 Figure 3 — System Structure of AutoHaul
This figure shows the system architecture of AutoHaul and segments categorized into operation centre, train-borne, communications, signalling, and
asset protection.

Global
Wayside emergency
systems Wayside attendant Train control Train Maintenance Software
button for ATH supervisor controller coordinator installer
Wayside equipment
Signals, points, track
Train planning
circuits
system, ETG

Asset protection
devices (DED, HBD, Train control
SFD, HWD) Operations
subsystem (TCS)
centre
systems

Level crossing
MLK Vital safety server Train-borne
interlocking (VSS) subsystem systems
CCTV office
Level crossing VICS (VICS_CC)

Communications
Level crossing infrastructure
auxiliary system subsystem
Train-borne:
(LXAS) (COMMS)
signalling system
Video image
capture system
DSE on-board
ATP (VICS_OB)
ATO_C

BSDS CDS
Tower controller Maintenance
Datalogger
ATO

Driver Train-borne: Locomotive control subsystem Locomotive interface

locomotive (LCS) including ECP and DLC gateway (LIG) Maintainer
system

Locomotive
interface

ATH: AutoHaul ETG: electronic train graph DED: dragging equipment detector HBD: hot box detector SFD: stream flow detector HWD: hot wheel detector
TCS: Train control system MLK: microlock VSS: vital safety server VICS: video image capture system CCTV: closed-circuit television LXAS: level crossing auxiliary system
CC: control centre COMMS: communication subsystem OB: on-board ATP: automatic train protection DSE: driving strategy engine ATO: automatic train operation
ATO_C: ATO controller BSDS: banker separation detection system CDS: collision detection system ECP: electronically controlled pneumatic DLC: direct locomotive control
LCS: locomotive control subsystem LIG: locomotive interface gateway

train-borne segment, and signalling and asset protection (1) Obstruction detection system
segment (see Figure 3). Laser based system detects objects within level crossing
boundaries
(2) Closed-circuit television (CCTV)
4. Level Crossing Protection Provides surveillance coverage of level crossing and
record footage via digital video recorder. Manual and auto-
Although the train runs through the remote Australian mated retrieval/viewing by authorized Operations Center
bushland there are some active level crossings along the personnel.
railway line that intersect with major highways or public (3) Streetlighting
roads. There are 42 active crossings on the Rio Tinto net- Provides consistent lighting for level crossing. Turns on
work which are protected by the level crossing auxiliary each time the crossing operates, is obstructed, or when staff
system, This can be broken down into three components. view the live feed from CCTV.

78.
 FEATURED ARTICLES

conducted by Hitachi’s railway business division up to now

into one standard companywide analytics platform.
5. Remote Locomotive Monitoring
6. 2
Integrated Operations and Optimization by Offering
A dedicated team of asset health evaluators and train con-
Fully Integrated Solutions
trollers monitor the health and status of wayside and train-
borne equipment from the Operations Center in Perth. The pit to port concept for mining offers a shift from the
Rio Tinto has about 221 locomotives equipped with the run when ready operating models to looking at demand
AutoHaul technology, for every locomotive that is on the and availability of dumpers and stockpiles; this will bring
network locomotive health status, alarms, and videos (col- efficiencies to the entire supply chain process. The solutions
lision detection events, level crossing obstruction events) highlighted below can be explored as steps to achieve this.
are monitored. Activities that were previously performed (1) Integrated traffic management system (TMS), which
by the driver while on board the train are now managed leverages off Hitachi train scheduling technology and
by the Rolling Stock Asset Health Evaluator Team, which interfaces with the port and mine, scheduling utilizing key
remotely monitors the onboard status of assigned trains inputs such as client demand on the type of iron ore blend,
throughout their journey and process all onboard alarms and shipping schedule, and ore stockpiles at different mines to
notifications raising maintenance request where required. client demand to create a daily, weekly operating schedule;
The AutoHaul system is designed to automatically stop (2) Train pacing and moving block functionality to improve
the train for: throughput, conserve fuel, and reduce wear on the train
(1) Asset alarms from wayside equipment like dragging while implementing operating schedule; and,
equipment detector or hot bearing detectors; (2) critical (3) Improvements to driving strategy engine (DSE) to han-
wayside and onboard equipment failure; (3) obstruction dle different train configurations, for example, two or three
events from level crossing; and (4) collision events detected head end leads, distributed power arrangement and different
on the train. wagon weights depending on iron ore blend and type.
6. 3
Connectivity
6. Further Opportunities
The Internet of Things (IoT) has the potential to expand
This section introduces the future vision of Hitachi’s rail- transportation capacity, improving maintenance and yard
way business division including establishing autonomous operation. This, combined with long term evolution (LTE*2)
operation system. and 5th generation (5G) technology provides an increase
in bandwidth needed for powerful visualization tools that
6. 1
can process information in real time for business-critical
Digital Value Proposition
decisions. Establishing relationships with market leaders
Using the power of analytics, it is possible to visualize the in communications, Hitachi can expedite to the advanced
vast amount of operational data available to identify trends market solutions to gain market share.
and predict issues benefiting maintenance and operations. In remote regions like the Pilbara, these technologies can
Maintenance strategy can shift from corrective mainte- also unlock other opportunities to introduce functionality
nance to preventive maintenance in the short term and such as:
ultimately to predictive and prescriptive maintenance. (1) Tablet-based human machine interfaces (HMIs)
Autonomous operations bring repetition, the advantage offering a safety integrity level (SIL) rated alternative to
of every locomotive driving the same way through the the current paper-and-voice-based method to take track
network this makes the wear and tear patterns on locomo- possession;
tive and trackside repeatable and predictable, which can (2) Tablet-based HMIs can also be used on track machines
be mapped to benefit in preventative maintenance tactics. and high-rail vehicles to integrate them into the AutoHaul
Longer-term machine learning through smart algorithms system architecture and improve safety;
can highlight potential problems before they become seri- (3) Minimize the installation on wayside infrastructure for
ous issues that result in taking out assets for a prolonged new stations that localize maintenance teams to central
period of time. This makes it possible to avoid a situation locations, reducing theft, power consumption, and driver
where assets such as vehicles and railroad tracks are unus- fatigue.
able for a long period of time.
Many synergies can be expected by sharing the use cases
*2 LTE is a registered trademark of the European Telecommunications Standards Institute
across the freight, metro, regional, and high-speed projects (ETSI).

Hitachi Review Vol. 69, No. 6 790–791 79.

 Authors
Mazahir Yusuf
7. Conclusions Hitachi Rail STS S.p.A. Current work and research:
Coordinating the engineering activities on the
AutoHaul System Enhancement Project. Society
This paper has presented an overview of the development memberships: Engineers Australia.
of Rio Tinto’s heavy haul freight transportation system,
AutoHaul, and the future prospects of Hitachi’s railway
business division. Anthony MacDonald, Ph.D.
The advancements in freight technology through imple- Hitachi Rail STS S.p.A. Current work and research:
Development of software techniques and design of
menting the strategies and developing applications dis- integrated rail systems solutions for Hitachi Rail STS.
cussed will add value to heavy haul clientele, but these
can also benefit other platforms like regional and inter-
city operations, as some of the challenges experienced are
similar. These initiatives will lead to positive impacts not Roslyn Stuart
only for the client, but also on United Nation’s Sustainable Hitachi Rail STS S.p.A. Current work and research:
Business development and bidding on projects in
Development Goals (SDGs), which have been committed Australia and Malaysia.
to by Hitachi.

References Hiroko Miyazaki

Digital Rail Solution Department, Railway Systems
1) Ansaldo STS, “Ansaldo STS Signs Framework Agreement Business Unit, Hitachi, Ltd. Current work and
with Rio Tinto Iron Ore in Australia” (Nov. 2010), http:// research: Business development and engineering of
sts.hitachirail.com/en/press-releases/ansaldo-sts-signs- digital solutions for railway systems.
framework-agreement-rio-tinto-iron-ore-australia
2) Ansaldo STS, “Ansaldo STS a Hitachi Group Company
Proud its Technology Has Enabled Rio Tinto to Launch
the World’s First Autonomous Heavy Freight Rail
Operation” (Jul. 2018), http://sts.hitachirail.com/en/news/
rio-tinto-launches-autonomous-heavy-freight-rail-operation
3) Rio Tinto, “Successful Rollout of AutoHaul Positions
WA as Global Rail Technology Leader” (Jun. 2019),
https://www.riotinto.com/en/news/releases/
AutoHaul-successfully-deployed
4) The Australian Financial Review, “Rio Confident AI Will
Unlock More Gains in WA Iron Ore Rail” (Nov. 2019), https://
www.afr.com/companies/mining/rio-confident-ai-will-
unlock-more-gains-in-wa-iron-ore-rail-20191128-p53evn

80.