HIRSCHMANN NETWORK SYSTEMS

.
.
.
.
.
.
.
.
.
.

Hirschmann Network
Systems
.

.

.

.

.

.

.

Distributed
Communication
Architecture
Industrial networking solutions with a

future

.

.

.


Hirschmann Boiler plate.....................................................................................................4

Section 1.............................................................................................................................5
Industrial Communications Systems...............................................................................5
Introduction.........................................................................................................................5
Hirschmann DCA - a strategy for the next millennium ....................................................................5
Strategic direction..............................................................................................................................5
Why do you need a network architecture?.........................................................................................6
Flexibility for the future.....................................................................................................................6
The lack of a single transparent automation and control network.................................................6
The network is a long-term major asset.........................................................................................6

Principles for industrial network architecture..................................................................7
Industry trends....................................................................................................................7
The need for high performance Industrial networks .......................................................9
Distributed Communication Architecture – DCA..........................................................10
A statement of direction for industrial networks..............................................................................10
Real time ...................................................................................................................................10
Migration ...................................................................................................................................11
Topology & Resilience................................................................................................................11
Management ...............................................................................................................................11

Performance ................................................................................................................................11
Cost. .........................................................................................................................................11
A blueprint for future industrial network growth.............................................................................11

Section 2...........................................................................................................................12
An industrial networking architecture for the next millennium................................12
The future of automation .................................................................................................12
The Vision .........................................................................................................................13
Criteria for network evaluation........................................................................................16
REAL-TIME....................................................................................................................................17
Ethernet............................................................................................................................................17
Legacy Fieldbus...............................................................................................................................17
MIGRATION...................................................................................................................................17
Ethernet............................................................................................................................................17
Legacy Fieldbus...............................................................................................................................17
TOPOLOGY & RESILIENCE........................................................................................................17
Ethernet............................................................................................................................................17
Legacy Fieldbus...............................................................................................................................17
Continued….....................................................................................................................................17
Ethernet............................................................................................................................................17
Legacy Fieldbus...............................................................................................................................17
MANAGEMENT.............................................................................................................................18
Ethernet............................................................................................................................................18
Legacy Fieldbus...............................................................................................................................18
PERFORMANCE............................................................................................................................18
Ethernet............................................................................................................................................18
Legacy Fieldbus...............................................................................................................................18
COST................................................................................................................................................18
Ethernet............................................................................................................................................18
Legacy Fieldbus...............................................................................................................................18

Section 3...........................................................................................................................19


The Ethernet Evolution..................................................................................................19
A brief history....................................................................................................................21
The road to deterministic Ethernet..................................................................................21
Ethernet developments over the past decade...................................................................................22

Evolving standards............................................................................................................23

Section 4...........................................................................................................................25
A time for change............................................................................................................25
Market dynamics...............................................................................................................25
Vendor opportunities........................................................................................................26

Section 5...........................................................................................................................28
Hirschmann’s DCA........................................................................................................28
A blueprint for future industrial growth.........................................................................28
The Hirschmann Ethernet Fieldbus Approach.................................................................................29
Real-time......................................................................................................................................30
Migration. ..................................................................................................................................30
Topology & Resilience. ...........................................................................................................31
Management. ..............................................................................................................................32
Performance ................................................................................................................................33
Cost. .........................................................................................................................................33
Benefits............................................................................................................................................33
What Hirschmann offers..................................................................................................................33
Summary..........................................................................................................................................34

2


.
.
.
.
.
.
.
.
.
.

Hirschmann Network Systems
Communications Strategy
Hirschmann's long-term communications strategy is based around the complementary
strands of industrial automation & control communication and enterprise-wide
communications, managed by a common management application, HiVision.
The Distributed Communication Architecture (DCA) describes a robust standards-based
Ethernet solution for all levels of the industrial automation and control environment,
managing and handling information from instruments and sensors to control devices
which intercommunicate with plant computer equipment.
DCA can be deployed throughout the
wide spectrum of industrial
applications. Factory automation,
traffic management and process
control are typical environments
where Hirschmann’s industrial
network solutions are being used.

With intranet/Internet access to the
control network managers are able to
view the shopfloor, data and activities
easily and cost-effectively.

Industrial networks need to provide two views of the factory or process - a view of
operations and a view of configuration/management/diagnostics. Both require traffic
management capabilities in the network to prioritize traffic and minimize congestion,
which DCA provides.
The Scalable Ethernet Architecture (SEA) is a strategic framework for scalable Ethernet
throughout the enterprise, from the workgroup to the enterprise backbone, comprising
advanced network devices and management software.

3


Hirschmann Boiler plate
Hirschmann Network Systems, a division of Richard Hirschmann GmbH & co, is the
leading manufacturer of robust system solutions, designed specifically for industrial
networking requirements. Part of Rheinmetall Elektronik AG, the highly successful
German industrial conglomerate, Hirschmann Network Systems have ambitious growth
plans and aim to become the number one supplier of industrial strength networking
solutions within the next
three years. With a broad
spectrum of products,
Hirschmann provides a
complete range of Ethernet
solutions for industrial and
corporate end users.

Customers come from all enterprises, industrial and public sectors, including chemical
and automotive industries, finance and banking, local government, education, the media
and health care. Hirschmann has performed particularly well in harsh industrial
environments where the emphasis is placed on “super-resilient”, deterministic networks.
The industrial product portfolio, IndustrialLine, developed specifically for the challenging
conditions of the industrial world, include maintenance free, long-lived, standards
compliant products that are easily installed within a plug and play architecture. Consisting
of hubs, concentrators and switches,
the IndustrialLine includes four
product families: ASGE, MC, MR
and the second generation DIN Rail
family of products all designed to
address the specific requirements of
mission-critical industrial
networking.

Steeped in a tradition of technological innovation, the first milestone for the company was
back in 1984 with the installation of the world’s first Ethernet network, employing Fibre
Optics at the University of Stuttgart. Today, Hirschmann has 100,000 hubs and switches
installed in over 15,000 networks world-wide while domestically; Hirschmann is the
prime network supplier to 150 of Germany’s Top 500 companies. Hirschmann continues
to develop innovative, high quality network systems with 15% of its annual revenues
invested back into R&D. Hirschmann is ISO9001 certified and belongs to all the
predominant standardisation bodies. These include the IEEE, Gigabit Ethernet Alliance
and ATM Forum, Open DeviceNET Vendor Association (ODVA), Profibus Trade
Organisation (PTO) and the Fieldbus Foundation.

4

.
.
.
.
.
.
.
.
.
.

Section 1
Industrial Communications Systems
Networks exist to support the needs of the factory and are the lifeblood of the
manufacturing process. However, it seems all this transferring data around between the
different layers of the current factory floor network is becoming too complex.
Hirschmann solves this dilemma. Instead of viewing factory networks as independent
layers, they are viewed as a single resource for data streams prioritised by application
needs. By viewing factory traffic as layered data streams, it is possible to forward data
using a set of rules that applies to all layers. Instead of compromising between the
different capabilities of the different layers of today's factory network, managers can use
them fully.

Introduction
Most factory floor networks are not ready to take manufacturing into the next millenium.
The DCA product line from Hirschmann provides manufacturers with a practical highperformance answer with the ability to operate distributed high-bandwidth networks,
delivering unmatched performance through sophisticated robust design

Hirschmann DCA - a strategy for the next millennium
The industry is dominated by legacy fieldbus solutions. So-called fast control networks

generally operate at a meagre 1 or 2Mbps and lack the ability to scale to multi-megabit
speeds and support thousands of devices. Newer fieldbuses like 12Mbps Profibus promise
higher performance, but with an accompanying expensive price-tag. Foundation Fieldbus
are now committed to using 100Mbps Fast Ethernet for the long awaited H2 specification.
Of these alternatives, it is only Fieldbus Foundation with the H2 standard that has the
potential to provide an optimal solution for Industrial automation networks.
This is the market opportunity targeted by Hirschmann's Distributed Communication
Architecture. Designed to meet the demands of the most mission-critical application,
DCA is optimised to deliver the deterministic performance, scalability and high resilience
required by these applications at price-points far below those of today's fieldbus solution.
Hirschmann's Distributed Communication Architecture describes a control network
strategy for the next millennium.

Strategic direction
Simply, Hirschmann's DCA network architecture defines the strategic direction for its
next generation Ethernet fieldbus products - IndustrialLine. The combination of new
demands on the factory floor network and the emergence of the intranet/Internet
technologies has pushed current-generation fieldbus designs to their architectural limits.
Although elegantly simple in concept, DCA is a radical rethinking of the control network
architecture - and also defines the strategic direction for the development of the
Hirschmann IndustrialLine products. The DCA architecture is the means by which
5


Hirschmann will deliver high performance and guaranteed quality of service for real-time
processes as well as easy, low-cost deployment, thanks to its compatibility with legacy
fieldbus solutions.

Why do you need a network architecture?
Users are going to be spending large amounts of money on new automation and control

networks to meet the forthcoming bandwidth and performance crisis, so it makes sense to
do it right first time. A well thought-out network architecture outlines the solution to this
crisis and gives customers confidence about a vendors capability to answer both current
and future needs.

Flexibility for the future
As the automation and control infrastructure changes over time, the network architecture
must incorporate the necessary flexibility to accommodate evolving user needs. Investing
in the network today will buy flexibility for tomorrow.

The lack of a single transparent automation and control network
The past lack of an appropriate automation and control network architecture coupled with
the lack of standardisation of vendor offerings has prevented the rapid development of
new products and new vendor services. The subsequent lack of competitiveness (or
dominance of any single vendor-driven set of "standards") and the complexity of current
three-tier control networks has opened a new window of opportunity for vendors who
want to embrace a new architecture.
The diagram below shows how and when Ethernet is going to push down from the
information level all the way to intelligent devices at the instrumentation level.

The network is a long-term major asset
For users, the deployment of an automation and control network and related equipment is
a major expense and as a long-term major corporate asset and utility, a coherent network
architecture justifies the spending of funds.
Network architecture identifies the major components of a network and how they relate to
one another. Since it is strategic in definition, individual components or devices may not
be currently available, but available in a time-scale of about 18 months. In essence, it
defines the ideal state of an actual implementation of a network. However, an architecture
does not specify the exact sizing and placement of its components.
6

Principles for industrial network architecture
Although hardware and software implementation differs, the underlying standards for
open, production management systems are the same as can be found in today’s business
systems. That means freedom from the expense of maintaining specialized, one-of-a-kind
systems to run their plants. Further, open systems unchain live manufacturing data,
enabling companies to distribute it freely across enterprise networks in real-time to people
who can use it to make a whole company run more effectively.

Changing manufacturing practices are leading towards a new industrial automation and
control infrastructure. As firms move into the global marketplace and implement
advanced production processes, new technologies - such as Internet, wireless
communications, graphical client/server applications, smart devices and decision support
systems - are being deployed to reduce costs and streamline operations.
However, these new tools and business processes create significant data distribution
problems from the device level to the back office. Companies employing the latest
automation and control techniques can expect a steep rise in bandwidth requirements,
along with multiple challenges as they embrace technology to improve vendors' and
customers' role in production.

Emerging production processes, integrated systems and control/communications
technology offer significant competitive advantages. For many years, the drive in
manufacturing has been towards streamlined operations, improved response time to
production schedule changes and the use of electronics to price and fill orders.

Industry trends
The Internet, and its associated technologies, has radically changed the way people go
about their business today. It has improved communications throughout society and is
now ubiquitous on a global scale. During the 1990’s the main user of the Internet has

been people as they provide the intelligence to filter and sort the fast amounts of material
available into useable information. This model is changing. In the world of office
automation Internet technology has been designed into the devices that support the
business and its infrastructure. Example of this evolutionary process can be seen in
products as diverse as photocopiers and printers to LAN routers and voice PBXs. So why

7


is this happening? The answer is simple - it makes sense! Giving intelligent devices the
ability to communicate with the outside world is a good thing.
In the case of the printer & copier automated ordering of consumables such as paper or
toner, either to the office administrator or the supplier by email both saves time, money
and increases availability of the device.
As for the PBX, the ability for a device to inform the maintenance company when
tolerances are exceeded and things start to go wrong, rather than wait for a complete
system failure, saves time and money for all concerned. The additional benefit is that the
technology differentiates the supplier through improved customer service & support.
This value proposition, “to save time and money whilst offering increased service and
support” has great worth in Industrial application where vast sums of money can be lost
in a relatively short time when production or processes are halted.
For the process and manufacturing industries, this is the year of change and a shift to new
technologies. Underpinning all technological trends is the move towards open, transparent
commercial installations based on intranet/Internet and away from legacy, vendor driven
systems.
Every part of the process control and automation industry - from embedded systems to the
Fieldbus Foundation - has recognised the importance of Ethernet and TCP/IP. Ethernet
has become the dominant network technology at the controller supervisory level. Every
Controller, PLC and DCS vendor has an Ethernet interface and it is now moving
downwards towards device and the I/O level.

8


The need for high performance Industrial networks
Adding these new processes, systems and technologies to today's automation and control
communication infrastructure will stress it unbearably. Bottlenecks caused by, typically
three, discrete networks (Plant, Control & Device) will need to be removed before
networks become a transparent and plant wide utility.

Over the past five years there have been many enhancements to the Ethernet standards,
especially in areas of determinism, speed and prioritisation. There is no longer any reason
why Ethernet cannot be used to build deterministic fieldbus solutions that are costeffective and open. Since Ethernet is already the network choice for business computing,
its presence at the control level will make
sensor to boardroom integration a reality
rather than a goal for manufacturers.

With the physical bottlenecks removed raw transmission speed needs to be increased and
management policies implemented to allow the various traffic types to be prioritised
according to needs.
The initial impact of adding new, bandwidth hungry applications will be on factory floor
network, followed by WANs, should a manufacturer want to make key manufacturing
data available to customers and other partners in its supply chain.

9


Distributing manufacturing data is also a bandwidth intensive proposition. Over the next
four years, manufacturing plant information generated by DCS equipment is expected to
increase by 20 or 30 times the current level. Similarly, a 10 or 20 times increase is

expected in PLC equipment collecting information from the factory floor.
Distributed control systems (DCS), Controllers and programmable logic controllers
(PLCs) also eat up bandwidth. These enabling technologies facilitate smart sensors and
devices on the factory floor. Smart sensors mounted on process equipment are now
capable of network connectivity throughout the factory; and each sensor being
individually addressable and intelligent.

Distributed Communication Architecture – DCA
The Hirschmann Distributed Communication Architecture provides suppliers and end
users with:



a statement of direction for industrial networks



a blueprint for future industrial network growth

A statement of direction for industrial networks
To meet the demands of the next generation of automation and control system, the
network will require a new architecture comprising six key dimensions: real time,
migration from legacy systems, resilience, management, performance and cost.

Real time
Whether you are designing a small, medium or large control network, Hirschmann
networks are designed to grow as end user needs grow and to meet the needs of higher
bandwidth real-time applications smoothly. With Hirschmann's Distributed
Communication Architecture, its policy-based QoS makes sure high-priority traffic for
certain messages always gets through.

10


Migration
Hirschmann integrates legacy devices, instrumentation and I/O through gateways
supporting existing control and device networks. These gateways also aid smooth
migration of installed control networks to Ethernet.

Topology & Resilience
Ethernet and TCP/IP based networks are inherently scalable by design. The Internet, with
its millions of end stations is testament to this. On a local level, single mode fibre optics
can move data at rates of 10, 100 or 1000Mbps over distances exceeding 20km with a
single hop. Multiple hops and differing topologies (bus, star or ring) extend this even
further. Created from the start as mission-critical products, Hirschmann's IndustrialLine
offers no single point of failure in a network, either physically or logically

Management
Hirschmann's Distributed Communication Architecture (DCA) offers comprehensive
management capabilities via Web browser, SNMP and priority-based VLANs.

Performance
Because of DCA's scalability, Hirschmann can give all level of the control and
automation network the bandwidth it needs at the level of the network which makes most
sense. Ethernet transmission speeds from 10Mbps to 1000Mbps are all fully standardised.

Cost.
Today, nothing can compare with Ethernet as the lowest cost implementation for a control
network. The ability to take advantage of the existing support infrastructure for Ethernet
is a major benefit to suppliers and dramatically reduces the total cost of ownership. Cost

is also a factor from a development perspective, TCP/IP communications software and the
underlying ASIC chips are commodity, mass market items and priced accordingly.

A blueprint for future industrial network growth


Greater openness/interoperability with other devices, management software
and control platforms. Hirschmann DCA provides an open communication
architecture compared to legacy control networking and connectivity.
Vendors want openness to reduce client software expense and increase
access to devices and other products.



Reduced dependence on costly, highly skilled field installation and support
functions - through automatic Internet Web connection and services.



Greater partnership with end users offering open independent solutions with
superior support services.

11


Section 2
An industrial networking architecture for
the next millennium
The demand for ‘open’ industrial communication systems is being driven by end users'
desire to move away from older, centralised plant control strategies to distributed control

in the field. End users want an enabling technology that provides true device interoperability, enhanced field-level control, simplified maintenance and reduced installation
costs. The only network architecture capable of delivering against these requirements will
offer deterministic high performance, be standards based and non-vendor specific.

The future of automation
New approaches to process and manufacturing automation, which will have a tremendous
impact on the design of control networks, include:


Manufacturing execution systems (MES). Computer-based information and
command for managing production resources, processes, costs, labour, data
collection, documentation, and work-in-progress, etc.



Data collection. New technologies for traditional functions such as time and
attendance recording, labour reporting, and materials tracking.



Computerised maintenance management systems (CMMS). Graphical views
of process equipment and processes accessed by multiple users to pinpoint
faults and failures, order design fixes and accelerate repairs and changes.



CCTV for monitoring



Data warehouses.



Quality information systems. For tracking compliance with ISO9000 and
industry-specific benchmarks.



Decision support systems.



Product data management.



Electronic forms. Data support for control and automation processes.

These new applications will demand higher bandwidth than ever before from plant,
control and fieldbus networks. This increased bandwidth is simply not available from
low speed fieldbus systems.

12


The Vision
Most plant data collection applications use a batch approach, where data is transmitted at
the end of the shift or other low usage times of day. New networking technologies will
change this model to real-time, with plant information being continually and

automatically collected and analysed - without operator intervention. Plant operations will
increasingly be directly connected in a client/server model to host computers and servers.
Controllers, PLCs and Enterprise Resource Planning (ERP) systems will be able to access
any sensor connected to the control and device network. The result will be better
information on manufacturing processes.

Imagine the impact of every shopfloor worker having the equivalent of a handheld,
possibly wireless network browser at his or her disposal. In real time, process operators

will be able to monitor and fine tune system performance, access plant information and
communicate directly with their production line managers. These operational online
continuous nodes will be another bandwidth consumer, raising traffic levels significantly.
And the network will not only supply information internally. Trends towards quickresponse, vendor-managed inventory and electronic commerce, are demanding that
manufacturing at the centre of the supply chain be brought online. Customers and
suppliers need to be able to look at all points in the supply chain, from initial order
placement to raw material consumption to assembly to shipment and delivery.
Decision support systems and data warehousing applications will soon be able to "mine"
massive amounts of data for correlation and trends that can lead to operational
improvements. With manufacturing equipment and personnel on the network, higher
management can have access to the operational data on the factory floor in unprecedented
detail.
The new factory floor network will also affect network capacity planning in the same way
as switched networks impacted on traditional shared LAN designs. In future, IT/network
managers will also need to be aware of developments on the factory floor generating
additional traffic which will impact office LANs and servers and, eventually, WAN
traffic. Another way of looking at this is of office users extending their reach towards the
factory floor. Table 1 shows how key business line functions correspond to plant
information sharing and data communication requirements.
Table 1

Function

Information sharing / Data requirements

Operations

Detail scheduling Sequencing, priorities, routings, shape, fit, setup,
alternative / overlapping / parallel operations, equipment loading, shift
patterns
Production units Flow, jobs, orders, batches, lots,work orders,

Dispatching

13


Process
management
Product tracking
and genealogy
Performance
analysis
Quality
management
Document control
Data collection/
acquisition

sequences, changes,events, schedules, controls, buffers
Monitor, control, correct, decision support, tracking, alarms, tolerances

Visibility, status, who is working on what components, suppliers, lots,
serial numbers, environments, alarms, rework steps, exceptions,
history, tracing, usage
Up-to-the-minute status, results, history, measurements, utilisation,
availability, cycle time, conformance to schedule, performance to
standards, parameters, reports
Analysis, measurement, collecting, quality control, identifying
problems, correlation, symptoms, actions, results, tracking, inspection
Forms, instructions, recipes, drawings, standard procedures,
programs, batch records, EC notices, as planned" and "as built"
Interfaces, links, production/parametric data, forms, scanned
transaction records, other collected data.

Complimentary technologies that support such network related business functions are
many and varied. They include:








smart sensors and fieldbus devices. Electrical panels and intelligent
circuitry in specialised local factory networks which enable multiple
performance measurements on production line equipment, operations
and materials.
high bandwidth devices. Ethernet interfaces, data collection terminals,
radiofrequency (RF) devices/transmitters and programmable
controllers.

traceability. RF tags, barcodes and smart cards.
client/server installations & thin clients
Internet access

As these technologies mature over the next couple of years, current bandwidth levels are
expected to hit their ceilings and new network solutions will be required. For control
systems, the immediate focus is the factory floor network.
Today, many devices are connected to a control network through proprietary serial
cabling and protocols. Information is then consolidated from the control network running
at speeds typically under 2Mbps. For this information to reach the corporate systems it
must cross the divide between the control network and the information network with its
links back into the enterprise office automation (OA) network. Typically this function is
carried out by a PC based gateway or HMI workstation. With interfaces to the proprietary
control network on the one side and the Ethernet/TCP/IP based information network on
the other, the gateway provides a route, albeit restricted, across the divide. In the majority
of cases today, the information network utilises 10Mbps shared Ethernet.
The first casualty of the information explosion will be the legacy fieldbus system with
2Mbps as its upper limit. Shared 10Mbps Ethernet will also be replaced initially by
10Mbps switched increasing to 100Mbps as need dictates. Ethernet will also extend its
reach, driving the technology closer to intelligent devices and remote I/O.
The Internet and modern networking designs will enable four major functions to be
radically improved.


Easier product installation for remote or local personnel using HTTP
server and browser technology set up, re-configuration and status
monitoring



Diagnostics/repair help find and solve problems with device or devices'
mission using SNMP, FTP, peer-to-peer or HTTP technology and
memory dumps to host for analysis, download programs to RAM or
flash memory
14




Use the HTTP server to gather a wealth of information from a device



Management reporting and network capacity planning

15


Criteria for network evaluation

16


REAL-TIME

Ethernet

Legacy Fieldbus

Application target

ALL

ALL

Determinism

YES

YES

Response time

4ms or less

5ms or less

Message size

ALL SIZES

LIMITED

MIGRATION
Geography

Ethernet

Legacy Fieldbus

WORLD-WIDE

REGIONAL

Backwards compatibility

HIGH

LOW

Degree of openness

HIGH

LOW

Interoperability

HIGH

LOW

Standardisation

IEEE 802.3

EN 50170 / Fieldbus
Foundation

Network Security

HIGH

HIGH

Protocol

TCP/IP

PROPRIETARY

TOPOLOGY &
RESILIENCE

Ethernet

Legacy Fieldbus

Automation level
Business

YES

NO

Control

YES

YES

Device

YES

YES

GATEWAY

YES

Bit-sensor

Physical Connectivity
Media

TP, FIBER, COAX, AUI,
WIRELESS

COPPER, FIBER

Devices connected

CONTROLLERS, FIELD
DEVICES, REMOTE I/O

CONTROLLERS, FIELD
DEVICES, REMOTE I/O

64000

500

1-256 (APPLICATION DEPENDANT)

48

Up to 40Km

Up to 20Km

Max no of nodes
Nodes per segment
Distance between nodes

Continued…

Ethernet
17

Legacy Fieldbus


Repeaters for longer
distances

YES

YES

Logical Connectivity
Communication modes

VARIED
YES

Internet/intranet

VARIED
NO

Resilience
Reliability

HIGH

HIGH

Scalability

HIGH

MEDIUM

Redundancy

YES

YES

Hot insertion of devices

YES

SOME

MANAGEMENT

Ethernet

Legacy Fieldbus

Plug'n'play support

YES

NO

Network topology

BUS, STAR, RING

BUS, STAR,RING

VLANs

YES

NO

HTTP / WWW

YES

NO

SNMP

YES

NO

1, 2, 3, 4, 5, 6, 7

1, 2, 7, User Layer

ISO Levels supported

PERFORMANCE

Ethernet

Data transfer rate

Legacy Fieldbus
LOW

10 / 100 / 1000 Mbps

Speed

HIGH

1-12 Mbps

HIGH

MEDIUM

Scalability

COST

Ethernet

Legacy Fieldbus

Cost

LOW

HIGH

Cost per connection

LOW

HIGH

Cost of Ownership

LOW

HIGH

MULTIPLE

SINGLE VENDOR

Sourcing

18


Section 3
The Ethernet Evolution
Advances in commercial networking technology have been coming so fast that it has
grown difficult for automation and controls suppliers to keep comparable elements of
their legacy systems abreast with state-of-the-art network developments. With the
networking industry offering dramatically increased bandwidth, commercially available
Ethernet equipment is now beyond what suppliers can hope to develop in-house in terms
of fieldbus and control networks.

For vendors, in-house development of computer hardware, operating systems and
networking elements, which can be purchased cost-effectively on the open market, exacts
a fierce toll. It is very expensive for suppliers to support 20 year old, proprietary systems.
Originally ignored by the Automation industry because of its perceived lack of
determinism and robustness, Ethernet has evolved into a technology which the
automation and control industry is swiftly adopting.
Ethernet TCP/IP is a widespread network technology, with users exceeding 100 million

world-wide. Ethernet PC boards sell for sub $30 compared to the $900 or more for a
control or device network PC board. In addition, the growing acceptance of Microsoft
Windows NT and its incorporation of Ethernet drivers into the operating system enhance
Ethernet as the backbone of high-speed control and device networks. Further, Windows
CE is being considered as an embedded operating system for devices and controllers,
leveraging Windows NT capabilities. Many PCs include an Ethernet network interface
card at little or no cost.

Ethernet TCP/IP also offers easy connection to the Internet, which is gradually filtering
its way into the world of industrial automation and control systems. Devices sitting on an
Ethernet TCP/IP network need only be assigned an IP address for Internet connectivity. In
addition, a complimentary Internet technology – Java, is already being used in
applications from automation and control suppliers.

19


To meet the demands of industrial control networks, Ethernet architecture must be based
on six main criteria: real-time capabilities, migration, topology & resilience, management,
performance and cost.

20


A brief history
The history of local area networking is relatively short - Ethernet was the first working
LAN. It was developed at the Xerox Palo Alto Research Park, beginning in 1973, by a
team headed by Dr Robert Metcalfe. Ethernet was first widely employed commercially to
network terminals to minicomputer systems - more specifically, to network Digital
Equipment Corp terminals to its VAX line of minis. Unix-based workstations and

scientific workstations were also connected by Ethernet early on.
The original published specifications were known as DIX (Digital, Intel, Xerox) Ethernet
Specifications Versions 1.0 and 2.0. The IEEE adopted, improved and modified the DIX
Version 2.0 specification. This became the IEEE 802.3 standard, which is equivalent to
the ISO 8802/3 standard. The first IEEE Ethernet standard was published in 1983,
defining what we know today as 10Base5 or thick Ethernet.
The earliest commercial network, Ethernet, used a bus - a single data path to which all
workstations attach and on
which all transmissions are
available to every workstation.
Only the workstation to which
the transmission is addressed
can actually read it, however. A
bus cable must be terminated at
both ends to present a specified
impedance to the network
workstations.

The road to deterministic Ethernet
How can only one computer at a time be allowed to transmit on the network? Access to
the network - the right to transmit - can be allocated in one of two ways: randomly or in a
deterministic order. In a random access method, any station can initiate a transmission at
any time - unless another station is already transmitting. In a deterministic access method,
each station must wait its turn to transmit.

Carrier-Sense Multiple Access / Collision Detect (CSMA/CD) is the random access
method used for bus arbitration within the original shared Ethernet standard of
IEEE802.3. CSMA/CD and random access are best suited where network traffic is
unpredictable and bursty, consisting of many short transmissions. Since industrial
networks are characterised by their deterministic nature, consistent low latencies and low

21


jitter, it is hardly surprising that the leading Automation vendors regarded early Ethernet
as unsuitable and developed their own networking solutions.
However, rapid developments in Ethernet switching technology in the early 1990s’ have
eliminated what were once barriers to the adoption of Ethernet as the control network of
choice. With its speed, robust performance, low cost of deployment and constantly
updated technology, Ethernet is a natural fit into the automation and control hierarchy.
Ethernet is typically used in manufacturing operations for communication both between
business system components and plant networks. Ethernet's capability to easily
communicate with multiple devices and manage the traffic to the information level of the
plant make it an ideal candidate for use at the control & device level.

Ethernet developments over the past decade
A new version of Ethernet, 10BaseT, appeared in the late 1980s. It uses twisted pair (TP)
wiring and is arranged in a star topology. Yet the network acts as a logical bus. That is to
say, signals transmitted by any workstation are available on the network to all
workstations. Only the station for which the transmission is destined can read it.
Ethernet switching arrived on the scene during 1992. The best analogy to switched
Ethernet is switched voice and the PBX voice switch. Ethernet switches also support
multiple simultaneous communications between many devices without collisions. Using
addressing information contained in each Ethernet frame a switch forwards data to a
switch port to where the destination equipment can be reached. The ability to switch an
Ethernet frame to a specific destination based upon information in the Ethernet frame
rather than broadcast the frame everywhere was the first step in making Ethernet
deterministic. Ethernet switching has since revolutionised the business of networking,
originally demanding a premium switched ports are now priced at a level where it makes
good sense to deploy them widely.
IEEE 802.3x brought with it standardised full-duplex operation and link based flow

control. Full duplex working in point-to-point mode, further overcame the issue of
determinism by giving a single station full wire rate connection, with no risk of data
collisions. It is
collisions caused by
two devices
attempting to transmit
at the same instant that
made Ethernet
unpredictable when
loaded. Where
contention is a
network issue, flow
control provides a
means whereby an
Ethernet switch can
indicate to
transmitting stations
that congestion exists
in the network and that they should pause transmission.
The next evolution was Fast Ethernet (IEEE 802.3u), which is nearly identical to 10Mbps
Ethernet. The packet length, packet format, error control and management information are
identical to 10BaseT, but the speed is increased by a factor of 10. It is implemented as a
star topology.
IEEE 802.3ac, supporting the work carried out by the 802.1 p & Q IEEE working groups,
added frame tagging for priority (8 levels) and VLAN identification which, when
combined with Ethernet switching, delivers deterministic performance.

22

Recently, Gigabit Ethernet (IEEE 802.3z) has been ratified, combining Fibre Channel
technology with Ethernet media access, running at 1000Mbps (1Gbps) over fibre optic
cabling.
Hirschmann has specifically addressed current concerns about Ethernet's lack of
determinism and redundancy using the standards described above and adding further,
complimentary features that are specific to building industrial networks. The result is
that Ethernet can now be adopted as the control & device network of choice. Several
control suppliers have begun to move towards open systems by taking advantage of new
technologies and incorporating them into their control systems. For example, Fisher
Rosemount's DeltaV system uses standard Ethernet as the control network between
workstations and controllers. Foxboro and Schneider Automation use Ethernet products
as the backbone of their control networks.

Evolving standards
Three phases in the evolution of the Ethernet fieldbus can be identified.
In the first phase, Ethernet replaces proprietary fieldbus connectivity and wiring as well as
proprietary fieldbus signaling implementations. Above Layer 1, all fieldbus protocols are

specific to the control network supplier. There are many examples of this type of
implementation such as Foxboro's Micro I/A Series which uses Ethernet for connectivity
between workstations and controllers, but retains a proprietary protocol to communicate
between controllers.

23


In the second phase, Layers 1 through 4 proprietary fieldbus protocols are replaced with
open networking protocols i.e. Ethernet TCP/IP. However, higher layers retain their
proprietary nature in order for suppliers to maintain differentiation. Examples of this
approach include Modbus which is underpinned with Ethernet TCP/IP but has added

application software in order to support the requirements of its field devices.

In the third phase, the trend towards a "standard fieldbus" becomes inevitable, driven by
user demand. The key example is the recent conversion of the Fieldbus Foundation for its
H2 fieldbus architecture to be based on Fast Ethernet. "The Foundation's Fast Ethernet
Phase 1
Phase 2
Phase 3
program is a direct response to the expressed needs and requirements of end users," said
Fieldbus Foundation President, John Pittman.

24