Introduction
Improving profitability in an increasingly competitive
manufacturing environment is a difficult challenge faced by most managers. Most manufacturers exist in a price-taker's
market. In such a market, the only
feasible way to improve profitability is to reduce operating costs, effectively
reducing the cost per unit of production.
Reducing the costs to maintain plant equipment still represents a
bountiful opportunity for improvement and cost reduction. Aggressive managers are recognizing that
aggressive maintenance management utilizing a PROACTIVE MAINTENANCE approach is
paying significant dividends. The
practice of reacting to breakdowns of critical production equipment is no
longer an option for the firm that wishes to continue healthy and profitable
operation. Utilizing machine monitoring
software in combination with existing IT infrastructure in PROACTIVE
applications is the best way to achieve long term cost reduction and profit
maximization.
Brief
History of Traditional Machine Maintenance
|
Ensuring production
uptime is the prime directive of the maintenance organization in a
manufacturing environment. In the
past, this has been accomplished by building in redundancies and excess
production capacity, or by following an aggressive schedule to rebuild or
overhaul critical systems. Both of
these approaches are inherently inefficient.
Redundant systems and excess capacity tie up scarce capital that could
otherwise be deployed in a producing activity. |
|
Scheduled overhauls and
rebuilds of critical systems and components offer managers peace of mind at a
very high cost. Intuitively, it seems
that rebuilding a machine on a schedule is the best way to avoid the costly
effects of a breakdown. However, when
one reviews this practice in more depth, it does have its’ risks. Most machines follow a probability of
failure pattern called the bathtub curve.
The bathtub curve simply displays a machines probability of failure
over time. It has three distinct
regions, the premature failure region, the random failure region and the wearout failure region (Figure 1). New and rebuilt systems enter their lives
in the premature failure region. The
probability of failure during this period is high because of all the
variables associated with manufacturing, machining, assembling and installing
a new or rebuilt system. Once past
this critical period, the system enters a period during which failures are
random and the probability of failure is statistically equal over time. At some point, all mechanical systems enter
a wearout period during which the probability of
failure increases. If a machine is
rebuilt on a schedule, it is removed from the random failure region where the
probability of failure is at its lowest, to the premature failure period
where the probability of failure is at its highest. The bottom line is that scheduling the
rebuild of a machine which follows the pattern of the traditional bathtub
curve actually increases the overall probability of a failure! This is a very expensive activity which
decreases the reliability of mechanical systems. Scheduled rebuilds and overhauls of
critical equipment is in conflict with the objective of extending the average
time between, and shortening the average length of, scheduled production down
periods for which most organizations today strive. |
Figure 1: Bathtub Curves Representing Generational Shifts in Maintenance Ideologies
Machine Maintenance
Evolves to Predictive
Advanced
maintenance organizations, recognizing the high cost of scheduled rebuilds,
have begun to utilize non-destructive testing techniques to identify failures
very early so appropriate repairs can be scheduled only when the machine
indicates that it is time for such an action (Table 1 & Figure 2). This approach to maintenance is called
predictive maintenance. Predictive maintenance offers numerous advantages
over a run-to-failure, or breakdown, approach to maintenance. And because maintenance activities are
scheduled in real time, according to machine conditions and requirements,
condition-based maintenance is far superior to traditional scheduled
maintenance. Costly unplanned downtime
is avoided and catastrophic chain reaction failures can be eliminated. With condition-based maintenance, overall
reliability is improved while the total cost of maintenance is reduced. Some of the technologies applied in these
predictive methods include vibration monitoring and analysis, wear debris
analysis and thermo graphic analysis. |
|||||||||||||||
Figure
2: Generational
Evolution of Maintenance Expectations The limitation of predictive condition-based maintenance lies in
the fact that it is failure oriented. Yes,
it is more effective than traditional approaches, but it leaves on the table
a considerable opportunity to improve reliability and uptime while reducing
costs. These benefits are available
only through PROACTIVE MAINTENANCE. Few
machines merely fail for no reason. The
majority of failures have one or more underlying root causes. Some of the root causes of mechanical
machine failure include: |
|||||||||||||||
|
|||||||||||||||
|
Controlling these causes of machine
wear and failure is the objective of PROACTIVE MAINTENANCE[1] (Figure 3).
|
|||||||||||||||
|
The human body
represents an excellent parallel to mechanical machinery to better understand
the opportunity which lies in PROACTIVE maintenance (Table 2). A breakdown, or run-to-failure approach to
maintenance is analogous to a heart attack or stroke. Waiting until this dire indication of
trouble in a human body or a machine results in the need to perform a quick
diagnosis and act immediately. There
is scarcely enough time to carefully acquire and analyze condition
information and make a thorough diagnosis of the situation. This leads to prescribed actions which have
a higher than normal probability of failure.
It is a situation which all physicians and maintenance mangers prefer
to avoid. |
|||||||||||||||
|
Table 2: Varying Maintenance Strategies
|
|||||||||||||||
|
In reviewing the human
body parallel to machine maintenance, the inefficiency of a scheduled
preventive maintenance program becomes clear.
No physician would suggest that critical body components be replaced
or rebuilt just because a certain age is reached. It seems equally illogical to prescribe an
overhaul or rebuild of a mechanical system based solely on a schedule,
without the assistance of machine condition data. |
|||||||||||||||
|
Most surgical
activities, such as heart surgery, are scheduled when non-destructive testing
information, such as an EKG, suggests that a problem is present. This information allows the physician to
acquire corroborating test information and diagnosis, and to schedule and
plan surgical activities under non-emergency circumstances, greatly enhancing
the probability of a successful outcome.
This is exactly the objective of predictive maintenance. By gathering machine condition information,
an effective diagnosis can be made, and activities scheduled logically and
with sufficient time to plan. |
|||||||||||||||
|
Most physicians today
recommend a PROACTIVE approach to human body maintenance. It is widely published that cholesterol and
high blood pressure are precursors to heart failure and other human ailments.
While the presence of either, or both,
does not in itself represent heart disease, they represent the underlying
root causes of heart failure, strokes, etc. By making an investment in controlling these
root causes today, an individual can reduce his risk of a failure later. Physicians recommend regular checks to
quantify the presence of these contaminants which are harmful to the human
body. When acceptable levels are
exceeded, prescriptive actions are taken to remedy the root cause condition,
not the component itself. This is
PROACTIVE health care. Machines can be
maintained in the very same way. By
regularly monitoring particle and moisture contamination (cholesterol to a
mechanical system), corrective action can be take to remedy the presence of
the contamination, eliminating the risk to machine reliability which they
represent. |
Computer-Managed Maintenance Systems
Definition
A
computer-managed maintenance system (CMMS) is an integrated set of computer
programs and data files used to manage the massive amounts of data generated by
maintenance, inventory control, and purchasing. It also incorporates work backlogs, schedules,
preventive maintenance plans and their schedules, labor utilization, and
maintenance cost distribution.[2]
Goal
“To
move all industrial maintenance environments to predictive maintenance - this
will allow for further progression along the maintenance evolutionary path into
the PROACTIVE MAINTENANCE arena.”[3]
As
is now widely accepted, maintenance is more an evolution than a goal. Despite the advances in technology there are
still a very large number of maintenance management departments that are
extremely reactive in nature. This is
identified by the proliferation of non-controlled stores systems, high levels
of reactionary or breakdown content, and by reading indicators such as MTBF
(Mean Time Between Failures) and MTTR (Mean Time To Repair).
A
maintenance workforce in a reactive state will have a very low MTBF of
equipment and an equally low MTTR. This
may be masked, if not measured and regularly reported, by the fact that machine
availability may still be at a high level.
This
indicator is informing the maintenance technician that there is a plant or
piece of machinery that is unreliable and breaks down often. It is also indicating that there is a talented
team of workers that are very good at fixing these breakdowns. The heroic culture that is fostered in these
sorts of situations can be the most difficult obstacle in the implementation of
a CMMS and realizing the gains of such a system.
Bottom
line, the achievement that is desired to be realized by implementing CMMS is
the advancement of the Maintenance Management workforce to the next level in
the maintenance process, the PROACTIVE stage (Figure 4). This is indicated generally by the fact that there
is greater control over stores systems, capacity scheduling techniques are
better utilized to more effectively manage human resources and planned backlogs
are at least two weeks out. Overall, this
results in much improved maintenance preparedness. Reliability and maintainability indicators will
improve with a CMMS system with MTBF rising and the MTTR measure, if it is
managed correctly, staying at the pre-implementation low level. This also gives a strong base for a move
onwards through the predictive stage of maintenance management through to the
World Class/continuous improvement stage.
Figure 4: Proactive Maintenance - Desired Move Resulting from CMMS
Benefits
of CMMS
Maintenance
Human Resources and Material Requirements
A measure that can be applied easily to a scenario where
CMMS is utilized to move an organization from a reactive maintenance
environment to a planned environment is a Planned versus Scheduled task
efficiency rate. A Planned / Scheduled
task will be 50% more efficient, in terms of duration and costs, than an
Unplanned / Unscheduled task. Although
this appears to be a large number, it is quite conservative and in some cases
the efficiency savings can be many times higher. Thus, the following calculation can be
applied with confidence:
(Past
years Unplanned/Unscheduled work (Dollars) – Unplannable
tasks) x 50%
Extending this calculation will result in an estimated
savings available through the use of a CMMS.
To do so, divide the above answer, pro rata, into the labor and
materials categories. This calculation provides
a very powerful and achievable savings indicator from a thorough implementation
of a good quality CMMS system. The
result is an easily achievable overall reduction in maintenance expenditure of
5%.
Machine
Availability
A thorough CMMS system implementation will also take into
account the KPI structure of the organization. With the amount of day-to-day data that will
now be available pinpointing problem areas and modifying processes, routines
and / or other factors will be more easily achievable also. As such, a reduction in the amount of breakdown
work, or increase in overall availability of plant and equipment of 5% is also
a very realistic and achievable goal.
Thus, the following calculation can be applied with confidence:
Past years breakdown downtime
(Dollars) x 5%
The above calculation can be used as a realistic measure to
measure the savings generated by a CMMS from the machine availability category
of maintenance cost.
Stores
Holdings
As the work content becomes more and more planned in nature,
the work of the stores department will become more and more predictive in
nature. This will allow for keeping a
lessened volume of parts and materials caused by a “we might need it” approach. As the stores holdings are reduced, the
logistical requirements to manage the stores function also decreases
dramatically, particularly that of the purchasing department and actual stores
management personnel.
A realistic and achievable reduction in stores holdings of
8% is realized through the use of a CMMS.
Following up on implementations with full stores reviews is recommended because
it will bring the focus more and more onto the analysis of what are critical
items and at what volume these items are needed given the improved operating
environment resulting from the CMMS.
Other
Areas
In addition to stores criticality reviews, there is a range
of suggested follow up exercises to a CMMS implementation. These include: 1) Maintenance strategy analysis and reviews
and 2) Root cause analysis reviews using the new reserves of accurate data. The focus created by the implementation raises
organizational awareness of the strategic importance of maintenance
improvement, and if managed correctly, leads to great initiatives that were
previously obscured by the needs of the company to “just keep it going”
Introduction to Web-Based Management of Real-World Assets
There are two sides to every
business: On one side, there are
databases, applications, and IT systems that help to run a business. On the other side, there are a host of “real
world” business assets—manufacturing machinery, security and environmental
systems, telecommunications equipment, storage tanks, and so on. In some cases, these assets are widely
dispersed, like in convenience stores, telecom sites, dispersed manufacturing
locations or other remote facilities. In
other cases, these assets may be right on the factory floor or in the on-site
data center.
In today’s enterprise, a chasm
exists between these two sides of business.
Information critical to your decision-making is hidden in real-world
assets and equipment. The means of
obtaining this information have been either non-existent or so complex and
expensive that the benefits do not warrant the implementation.
So how can this chasm between these
two sides of business be bridged—simply, affordably, and effectively? There is now a “plug-and-play” emerging
technology tool that uses standard Internet technologies and existing networked
IT infrastructure to connect databases and enterprise IT systems directly to
the essential line-of-business assets and equipment.
This is thought of as connecting the
business world to the real-world. Building this bridge allows organizations to
establish a vital link between the two sides of their business. As a result, companies are now improving
their overall operations by gathering key business intelligence from throughout
their entire enterprise and incorporating it into their business systems.
Web-Based Management of Real-World Assets
Uniting all corporate assets
underneath an IT platform is now possible through the melding of CMMS
technologies with older IT networking technologies. This idea is now being extended even further
due to quickly exploding internet technologies.
Although the overwhelming influence and impact of the Internet has been
widespread, one key arena that is still emerging is the effective use of
Internet technologies to monitor and control the non-IT types of assets and
systems that often form the heart of company operations (i.e. machining
equipment). While the potential
efficiency and productivity benefits of bringing these assets under the IT
network umbrella can be huge, a variety of implementation barriers have made it
very difficult to effectively bridge the gap between the cyber world and the
real world. In most cases, the
challenges of investing in or developing complex middleware-based control
structures and the IT staff costs associated with deploying and maintaining
such systems have represented virtually insurmountable barriers.
In spite of the barriers, the recent
convergence of Internet standards and functionality with proven and
long-established input/output (I/O) system technologies used for communicating
with and controlling production equipment has now opened the door for simple,
cost-effective methods for extending the Internet to encompass virtually all
real-world devices and assets. Leaders
in real-world I/O system solutions are forging the way with new-generation
hardware-based communication systems that can connect existing real-world
devices directly to Internet-based networks.
Web-Based Monitoring Barriers for Non-IT Assets
In most companies, the primary
operational activities that form the heart of the business are generally
dependent upon various real-world tangible operating assets that represent both
a major cost factor and a critical cornerstone in creating/maintaining the
corporate revenue stream. While in some
companies these core assets might primarily consist of easily networkable computing platforms, the overwhelming majority
of corporate assets are non-IT oriented systems and specialized equipment that
are not designed or readily adaptable for direct connection to computer
networks (i.e. machining equipment).
For example, in the airplane parts
manufacturing industry, these critical operating assets might consist of mills,
lathes, access controls, alarm systems, heating, ventilating, and air
conditioning (HVAC), utility management systems, lighting, paint guns,
autoclaves and ovens. In contrast, the core operating assets in communication
industries often include widely dispersed microwave towers, cellular/wireless
base stations, fiber-optic repeater stations, and other remote facilities. For utilities, petroleum, or chemical
companies, the real-world assets entail vital infrastructure elements such as
pumping stations, power-generating plants, distribution grids, pipelines, and
refineries.
Even for other companies that do not
necessarily have extensive specialized infrastructure requirements, the
real-world assets have not historically been designed or built for any form of
networked communications. Many of the
legacy systems were widely deployed in their respective industries long before
the relatively recent rise of computer networking technologies. Traditionally, the development of real-world
equipment has almost always focused more on maximizing the specific operational
requirements at hand, rather than incorporating ancillary functions such as
"ports and pipes" for networked communications.
The disparate design heritages
behind most of today's real-world operating assets have also made it
impractical to build in any sort of standards-based universal control
mechanisms for use across different types of equipment. In addition, even as such possibilities are
now emerging with the rise of the Internet, many of the existing infrastructure
assets represent millions to even billions of dollars of investment that cannot
be cost-effectively retrofitted or replaced to incorporate new communication
capabilities.
Monitor
and Control Opportunities Presented by the Web
As Internet technologies and
standards have rapidly developed over the past decade, it has become readily
apparent that Web-based control methodologies now represent a powerful
opportunity for extending efficient network-based management techniques to
encompass non-IT real-world assets.
The Internet's ubiquitous reach and
familiar Web browser interfaces make it a natural platform for implementing
remote equipment monitoring and control applications. Management applications software and user
interfaces can be developed using a wide array of Web development tools and
initial operator learning curves can be virtually eliminated by leveraging the
almost universally understood point-and-click simplicity of Web browsing. From a communications infrastructure
standpoint, the Internet's global reach, inter-networking flexibility, and
uniform communication mechanisms (such as TCP/IP and PPP) make it a highly
efficient and effective medium for deploying geographically dispersed control
systems.
In addition, the use of
Internet-based standards for communicating with and controlling real-world assets
also opens the door for IT departments to deliver a much higher level of
overall cross-integration between enterprise-wide strategic and operational
management systems. For example,
real-time data from sensors monitoring remote machining assets can be
automatically interfaced with overall maintenance planning applications,
service dispatch/scheduling systems, spare parts inventories, capital asset
databases, and strategic performance review and reporting systems. As a result, corporate-wide resources can be
more efficiently allocated and managed on a tactical basis while simultaneously
improving overall strategic planning and management with more complete and
timely information surrounding all critical company assets.
Barriers
to Maximizing Internet Opportunities
Clearly the Internet offers some
very powerful possibilities, once the data can be brought into the
corporate-wide network and integrated within the overall scope of IT-driven
productivity applications. However, the
primary difficulty with implementing any Internet-based asset management
architecture lies with creating the far-end connections between the edge of the
standards-based network and each of the disparate non-IT devices that need to
be managed.
Over the past few years, the
attempts to resolve this "last-yard" connectivity challenge have
generally focused upon replicating a far-end network node next to the remote
equipment by using some sort of PC-based controller running standalone
custom-written software. However, this
PC-centric approach carries with it a number of inherent shortcomings. For example, for most remote real-world
assets, deploying a full PC to provide a monitoring or control connection is an
overkill solution in terms of both cost and complexity. In many cases, the cost of the PC itself and
the associated software may be prohibitive and for most remote monitoring
applications, a standard PC configuration must be supplemented with
industrial-grade housing for harsh environments and specialized add-in
controller cards and sensors to capture machine-specific or process-specific
information.
In addition to the cost of the
hardware, most PC-centric remote monitoring or control applications also have
to rely on relatively complex software and middleware structures to process and
communicate the relevant data. In many
cases, this has necessitated the development of unique, one-off,
from-the-ground-up communication interfaces and controller designs for each
individual application, thereby failing to capture and leverage the inherent
efficiencies available from using standards-based networking mechanisms.
Another major concern with
PC-centric and software-intensive control architectures is the fact that they
are inherently more costly to maintain and support. The open-systems nature of PCs often makes it
difficult to establish and maintain uniform configuration control, especially
when attempting to deploy them by the thousands across many geographically
diverse remote monitoring and control sites. In addition, the familiar user interfaces
associated with every PC can leave them open to the potential for unauthorized
use or modification, leading to higher ongoing service costs. Since in virtually any IT organization the
highest cost factor and most precious resource is the finite availability of IT
staff time, the inherently higher support requirements for remote PCs,
specialized software, and middleware can quickly offset most of the potential
benefits from remote asset management.
Converging Proven Technologies and Capabilities
Leaders
in the advanced I/O system solutions for automating and controlling a wide
range of production, monitoring and other operational equipment, are taking the
innovative steps to Internet-enable proven product technologies. This convergence of proven hardware
controllers and Internet capabilities has resulted in a whole new breed of
flexible and efficient hardware-based, standards-compliant solutions for
managing real-world assets via the Internet.
Complete Hardware-Based Solutions
Unlike
other previous attempts at managing remote real-world assets, the new
hardware-based solutions provide a completely self-contained and virtually
maintenance-free bridge between any real-world asset and the existing
standards-based IT infrastructure (Figure 5).
Figure 5: Hardware-Based Web Computer Monitoring and Control
Wide Spectrum of Monitoring/Control Capabilities
Leveraging
proven I/O technologies, the new systems are able to quickly and easily capture
relevant operational and status data from virtually any electrical, mechanical,
or electronic real-world device. Device
information includes:
Serial
data-access controls, instruments, recorders, printers, barcode readers,
scanners. Analog signals-temperatures,
pressures, fluid levels, humidity, flow rates.
Digital
signals-on/off status, contact closures, alarm points, door sensors.
Standard SNMP and Web Browser-Based Management Mechanisms
Using
standards-based network communication mechanisms, the new systems then transmit
the device information across the network to any IT management application
using standard Simple Network Management Protocol (SNMP) mechanisms. The ability to communicate with the system
using a familiar Web browser and Web pages stored on a network server
streamlines the system setup process and enables quick configuration to the
required monitoring tasks. In effect,
SNMP traps are established for any of the real-world events, conditions, and
parameter limits to be monitored, and then the SNMP traps are automatically
sent to a designated host server when the triggering events occur, such as a tolerance
on a mill being exceeded or a temperature exceeding a preset value.
The
use of a standard Web browser interface for all setup and management functions
provides for complete configuration flexibility from a centralized management
location. The new self-contained systems
simply connect at the remote location to both the real-world device and the
network interface; after this, the system can be auto-discovered and completely
configured by centralized IT staff. This
remote configurability not only leverages maximum efficiency from scarce IT
staff resources while avoiding unnecessary deployment of maintenance personnel,
it also minimizes the risks of unauthorized configuration changes at the remote
location.
Implementation
Simplicity and Flexibility
From
a network connectivity standpoint, the new system architecture also provides
maximum flexibility for interfacing across a complete spectrum of real-world
deployment requirements. For example, in
addition to SNMP, the SNAP-IT systems also support other standard networking
and data communication protocols, including TCP/IP, UDP/IP, PPP, SMTP, HTTP,
HTML, and XML. The physical network connection
to the system can be via a wire-line Ethernet connection, analog modem, DSL
connection, or via a wireless link. By providing a variety of standards-based
networking connections on the upstream side and a wide range of configurable
data acquisition I/O alternatives on the downstream side, the emerging system
provides an ideal interface for monitoring and controlling real-world devices.
Real-World Application – Critical Process Monitoring
The new hardware-based systems can be effectively deployed to provide real-time monitoring and management of critical processes, such as machining processes, from which the failure to consistently maintain strict process parameters can have dire safety and financial consequences. For instance, the new system can be configured to monitor pressure levels, detect leaks, regulate flow rates, and monitor environmental conditions surrounding the installations. In addition to logging and reporting on remote process conditions via standard network management mechanisms, the new hardware-based system can be configured to automatically initiate predetermined safety measures such as shutting down machines.
Conclusion – The Bottom Line on Web-Based Monitoring and
Control
|