每日免费获取报告
1、每日微信群内分享5+最新重磅报告；
2、每日分享当日华尔街日报、金融时报；
3、每周分享经济学人
4、每月汇总500+份当月重磅报告
（增值服务）
扫一扫二维码
关注公号
回复：研究报告
加入“起点财经”微信群
Schneider Electric White Paper2
Asset Performance
The Evolution
of the
Maintenance
Function:
What’s in a
Name
Industrial maintenance has advanced considerably over the last two decades. Twenty
years ago, the predominant maintenance strategy involved reactive maintenance—waiting
for a piece of equipment, sometimes referred to as a plant physical asset, to break,
and then fxing it as quickly as possible. Results improved as maintenance engineers
developed more sophisticated preventive strategies, analyzing equipment to determine
the normal time-to-failure and scheduling maintenance to prevent failures before
they occurred. More recently, complex predictive maintenance strategies have been
developed, which involve directly measuring equipment conditions, such as vibration or
casing temperature, to forecast the probability of failure and then scheduling maintenance
procedures to fx the problem before it happens. Today, most industrial operations utilize
all three strategies.
As methodologies became more advanced, industrial maintenance communities evolved
the name associated with the maintenance function. Traditional maintenance management
shifted to asset management then to asset performance management, with each new
name intended to indicate the increased sophistication.
Although this naming evolution might seem appropriate for indicating increasing levels
of complexity, the focus on asset performance management might be both too narrow
and too broad when analyzing how to make industrial assets perform better: too broad
because the performance of industrial assets involves more than just maintenance
strategies, and too narrow because optimal performance of industrial assets should
include both transactional management and real-time control aspects.
The primary objective of industrial companies is to safely maximize proft from production,
and this requires a two-tier model, encompassing both asset performance management
and asset control. When combined with real-time information gathering, reported in
appropriate units, this approach creates a balance between the need to maximize
reliability and effciency goals with safety constraints.
The performance of industrial assets is a function of both asset maintenance and asset
operation. In most plants, maintaining the asset and operating the asset are performed by
completely different teams, with completely different—and often conficting—performance
measurements. The maintenance team is typically measured on the reliability of the asset,
while the operations team is often measured on the production throughput, or effciency,
of the asset. The problem is that reliability and effciency tend to be inverse functions; that
is, improving reliability typically involves reducing production throughput, and improving
production throughput typically means reducing reliability.
No wonder maintenance and operations teams have diffculty cooperating. They are
measured in a manner that penalizes one or the other, or both, for cooperating. This
industrial conundrum (fgure 1) must be solved if manufacturers truly want to maximize
asset performance across their industrial operations.
The primary objective of
industrial companies is to
safely maximize profit from
production
Effciency vs
Reliability:
The Industrial
Conundrum
Schneider Electric White Paper3
Asset Performance
Efficiency – Reliability Conundrum
Rel
iab
ility
Efficiency
The key to overcoming this confict is developing performance metrics for operations and
maintenance that encourage cooperative behaviors. Such a measurement system needs
to be indicative of the performance of the assets to the business. Every physical asset in a
plant exists to drive value, so performance indicators must measure the value each asset
contributes to the operation.
Since business leaders in most industrial operations measure value in fnancial terms,
basing asset performance on fnancial metrics typically provides the measurement system
that will encourage the maintenance and operations teams to cooperate toward a common
goal: maximizing the business value from each industrial asset and asset set. Fortunately,
developments in business performance measurement systems have resulted in real-time
accounting (RTA) systems that utilize plant-sensor data in combination with fnancial data
to model the business performance down to the asset level. These provide the necessary
fnancial contribution metrics.
Efficiency – Reliability Solution
Rel
iab
ility
Profita
bilityOper
ationa
l
Efficiency
Industrial companies are in the business of making money. Therefore, their operations
should be thought of as the “proft engines” of their business. Traditionally, industrial
business and operations functions were performed independently. Operations worked
to make products; business teams worked to account for the products made and sold.
Since the speed of business has continually increased over the past two decades, this
separation has become impractical and ineffective. Decisions being made second-by-
second on the plant foor have signifcant impact on the proftability of the business.
Those decisions, whether automatic or manual, must drive maximum proftability for the
Figure 1
Under the current model,
improving reliability tends
to decrease effciency, and
vice versa.
Figure 2
Using business metrics
provides a solution that
balances effciency and
reliability.
Real-Time
Control and
Revenue:
Igniting the
Proft Engine
Schneider Electric White Paper4
Asset Performance
business. However, developing a system that ensures these decisions are being made in
a coordinated, proft-generating manner is not a trivial task. The key to igniting the proft
engine of industrial operations is developing a control and management system that safely
maximizes operational proftability.
When designing the optimal control and management system, it is helpful to understand
how industrial operations have evolved over the last century. Real-time process and logic
control have been implemented, in one form or another, for the better part of the last 100
years. These systems were used to apply real-time controls to maximize the effciency of
industrial assets and asset sets. Over the years, considerable advancements have been
made as process control systems evolved from feedback control, to feedforward control,
to multivariable predictive control.
Process control operated under management systems designed to ensure effective
operation and follow appropriate production schedules. The control systems made
decisions in real time (within the time-constant of the process being controlled), while the
production management systems made decisions on human schedules, such as daily,
weekly, and monthly, commonly referred to as transactional decision-making.
As the sophistication of process management and control systems increased, the
effciency of industrial plants steadily improved, which in turn pushed the process
equipment harder and harder, until they reached their thresholds. Plant engineers quickly
learned that the solution was to implement safety control systems to ensure the thresholds
were not crossed. Additionally, as effciency increased and the equipment was pushed
harder, reliability of assets started to decline. Again, plant engineers learned that the
solution was to implement more advanced maintenance tech