Other than personnel-related costs,
maintenance is generally the most expensive
item in the laboratory operating budget.
While the size of this expenditure
makes it an attractive target for reduction,
all laboratory managers recognize the limitations
they face in achieving any real
savings. Foregoing maintenance and
repair is usually not an option—they are a
necessary cost of continued operation.
Thus, cost reduction initiatives must
extract operational and management efficiencies
from the maintenance system
without directly impacting instrument
reliability. Given the progress that has
been made in controlling the cost of laboratory
expendables and the effectiveness
of budget constraints in limiting discretionary
expenditures such as
travel, training, and conference
attendance, improvements in maintenance
management remain one of
the few fertile areas to realize additional
cost reductions through operational
efficiencies.
Evolution of maintenance philosophies
Throughout most of the twentieth
century, maintenance was primarily
reactive; instruments and equipment
were used until they broke
and then maintenance was called
to repair them. As equipment
became more complicated, the philosophy
of “run to failure” was augmented
by the concept of preventive
maintenance to delay the
failure. The central idea was to prevent
equipment from failing by replacing
key parts before they wore out. This
approach continues to be used to some
extent by virtually every laboratory.
While this idea seems eminently logical
and reasonable, data collected by a
Federal Aviation Administration
(FAA)/airline industry committee
chartered to study maintenance strategies
revealed surprising results.1 The
data showed that the common belief
that reliability declines with increasing
age is generally not true for complex
equipment and that scheduled maintenance
generally has little effect on the overall reliability. In fact, for many
instruments, there simply is no effective
form of scheduled maintenance.
Even more disconcerting was the realization
that preventive maintenance
might actually introduce additional risk
of failure. For example, the service technician
might accidentally damage the
affected or adjacent equipment in the
course of the inspection, repair, adjustment,
or installation of a replacement
part, or might install defective parts, or
incorrectly reassemble the equipment. It
was also found that equipment is more
likely to fail early in its life than later—an effect known as infant mortality.
This effect is familiar to most laboratory
managers in the case of computers,
where problems are more likely to surface
within the first few weeks of use
than in subsequent years. Thus, each
installation of new parts during preventive
maintenance reintroduces some
degree of risk of the infant mortality
effect. This does not imply that preventive
maintenance is ineffective or
should be discarded, but only that it
should be used judiciously within a
broader, more strategic approach that
considers the effectiveness of each task
to ensure that the benefits are commensurate
with the risk.
Realizing that it is virtually impossible
to prevent equipment failure, the
focus shifted from prevention to the
concept of preservation of function.1
This philosophy, known as reliability centered
maintenance (RCM),
accepts that equipment will always fail
but seeks ways to preserve function.
The basic principles are:
- Focus on preservation of system function
- Identify specific failure modes
- Rank importance of failure modes
- Identify effective means to mitigate the highest-ranking modes.
Under this system, the objective
of the preventive maintenance
program is to alleviate the consequences
of failure rather than prevent
the failure. Thus, if the consequence
of a particular failure
mode has no adverse effect on
safety, operations, environment,
or cost, there is no need for scheduled
maintenance. Due to the
resources required to identify possible
failure modes, the scope of
RCM programs is generally limited
to a small segment of instruments
and equipment deemed
truly business critical. A decision
tree defines the preferred maintenance
strategy1 to preserve function
for each of the likely failure
modes that includes options such
as run-to-failure, redundancy,
scheduled discard and replacement,
equipment redesign, or more
advanced maintenance techniques.
One of the advanced techniques used
in RCM is predictive maintenance.
This approach strives to use technology
to detect the onset of equipment
degradation and to address problems
as they are identified. It differs from
preventive maintenance in that needs
are based on the measured condition
of the equipment rather than on a predetermined
schedule. Thus, component
operational life and availability
can be extended, equipment downtime
for servicing is decreased, and maintenance labor and parts expense
are decreased. Common technologies
used in predictive maintenance are
vibration analysis, lubricant metals
analysis, and various on-line monitoring
sensors to indicate equipment
wear and progression toward failure.
Unfortunately, the high cost of test
equipment and expert resources
required to properly employ this technique
generally limit its use to large,
high-value mechanical equipment.
The popularity of total quality management
(TQM) during the 1980s and ’90s
extended into maintenance philosophies
with the introduction of total
productive maintenance (TPM).2 This
approach recognizes the importance
of the role of the operator and
teamwork in achieving and maintaining
the highest level of equipment
reliability. Proponents of this
philosophy believe that equipment
should have its lowest reliability on
the day that it is delivered and
should undergo continuous
improvement throughout its useful
life. Anyone who operates, maintains,
purchases or stores parts,
modifies, installs, programs, makes
decisions, assigns work, or otherwise
has a direct or indirect effect
on the reliability of an instrument
should be involved in its maintenance.
The philosophy embraces all
of the elements of RCM, predictive
maintenance, risk analysis, and
other advanced techniques, but
extends to include the softer teamwork,
attitude, and behavioral
issues common in TQM programs.
Management options
Many laboratories have internal
instrument technicians who provide
first-level support by performing most
repairs on chromatographs, ovens, and
other relatively simple equipment.
Instrument companies (OEMs) facilitate
this approach by offering low-cost
training courses on equipment maintenance
and repairs that provide sufficient
knowledge to perform tasks such
as rebuilding detectors or pumps, troubleshooting
flow problems, or
exchanging circuit boards. Some
OEMs also provide excellent call center
support to facilitate these in-house repairs by stepping the technician
through the diagnostic procedure.
However, the majority of an instrument
technician’s time is spent on
preventive maintenance and calibration
tasks; OEMs or independent service
providers (ISPs) are typically
used only to assist with overflow work
or to handle repairs or tasks beyond
the ability of the technician after an
initial assessment. In addition, service
contracts are purchased for certain
complex or potentially hazardous
equipment such as X-ray spectrometers,
in which the OEM is the sole service
provider. This operational model
can produce relatively low costs provided
there is sufficient work to keep the technician busy. Fully burdened
labor rates for internal resources are
typically less than OEM service technician
rates, and travel-related
expenses are avoided.
While the in-house maintenance
model appears to be a cost-effective
option, there are often hidden costs.
For example, in large companies, inefficiencies
in management of capital
inventory and coordination of multiple
service contracts can result in
higher costs and the loss of any potential
savings.3 Roles and responsibilities
shared with other departments such as
purchasing or central maintenance are
often ill defined, and the laboratory manager who has primary responsibility
typically lacks sufficient time to
properly manage this function. Laboratories
in regulated industries or
those that have achieved accreditation
to the ISO 17025 standard face
an additional administrative burden
in providing documented, auditable
records of all work performed. Reporting
requirements are significant and
tend to be neglected or degrade in
quality over time unless aggressively
audited and managed.
Over the past few years, several commercial
vendors have developed maintenance
services to introduce efficiencies
into this function, improve equipment
reliability, relieve much of the burden
of management, and bring a new
order to the entire system. Instrument
service contracts and point-of-need
repair services have been supplemented
with more complex and
comprehensive programs that
promise not only higher reliability
but also 15–25% lower costs. The
most sophisticated plans extend to
include often badly neglected areas
such as equipment inventory control
and disposition services. Some of the
common commercial options are
managed maintenance, multivendor
repair, and comprehensive or total
facility maintenance services.
On-site multivendor repair (MVR)
services are offered by ISPs and several
OEMs, where a single vendor
performs maintenance on all brands
of equipment. This service is similar
to the internal model except that
management and staffing of the function
is delegated to the commercial provider.
Laboratory labor rates for MVR instrument
technicians include the provider
overhead and are typically similar to or
greater than internal rates but still offer
savings over OEM contracts. As with in house
services, these providers typically
limit repairs to a relatively few types of
commodity equipment that constitute
the bulk of the laboratory’s capital
inventory. They may also maintain parts
inventories to speed repairs and provide
generic repair training for their service
technicians. Negotiated OEM service
contracts or demand services may stay in
place for the remainder of the equipment.
While managers appreciate the cost savings of this approach, scientists
responsible for the instruments are often
concerned about relinquishing control
of the quality of repair to the MVR vendor
and are reluctant to turn their instruments
over to technicians they view as
less qualified. The skill gap is a greater
concern with newer, more complex
equipment, where it is difficult for the
MVR providers to keep abreast of the
latest technology when they do not have
access to the OEM training resources.
For this reason, most laboratory managers
take a very cautious approach to
selecting this option.
The managed maintenance model continues
to rely on OEM repair services but
provides a single administrative point to
manage all contracts. By maintaining
accurate inventory records and aggregating
contracts, service providers are able
to guarantee cost savings while providing
additional value-added management services.
The primary concern with this
model is a potential decline in service
response time if the OEMs give preferential
treatment to their direct service contract
customers. The reality is that OEMs
are typically customer oriented and provide
good service levels regardless of the
maintenance funding mechanism. This
issue can also be addressed during the
purchase negotiations for any of these
services by identifying the truly mission critical
instruments and equipment and
designating them for guaranteed priority
service. The service provider can then
secure priority service for the designated
equipment. Fortunately, only a small
number of instruments are typically critical
to the operation of a laboratory; thus a
reasonable downtime is acceptable for
the rest in order to achieve the cost savings
and retain the quality relationship
with the OEM—the objective in this
model is to optimize performance rather
than maximize uptime at any cost. While
any instrument downtime is an inconvenience,
scientist productivity is typically
not impacted since they simply switch to
other equally important tasks until
repairs are completed. Since the managed
maintenance model does not compete
with the OEM, the customer continues
to benefit from factory-certified
technicians, access to parts and diagnostics,
application support, and continued
remote troubleshooting services that can
quickly and accurately identify and fix
the problem.
For those managers who prefer to
employ advanced maintenance
philosophies such as TPM, managed
maintenance is still an attractive
option for the coordination, administrative,
and reporting tasks. The
instrument and equipment operators
provide daily preventive maintenance
service to the instruments, while the
managed maintenance program provides
a disciplined and systematic
approach for low-cost supplemental
repair services to relieve both the
manager and staff from the administrative
burden. This model can be
more difficult to implement, but could
offer the highest reliability at the lowest
cost if successful.
Conclusion
Maintenance, a fundamental element of
laboratory operations, is expensive and
requires tedious administration to function
effectively. This makes it a prime candidate
for outsourcing to one of the new
commercial services. Full service vendors
with core competencies in asset management,
managed maintenance, validation
and compliance services, and disposition
services (e.g., the LIFECYCLE® program
[Thermo Electron Corp., Waltham,
MA]) are able to bundle service options to
deliver higher quality at guaranteed lower
costs while allowing the laboratory manager
to focus on laboratory operations and
business issues. In addition, a complete
suite of metrics and reports provide feedback
and assurance that the system is
functioning properly—a luxury that most
managers today do not enjoy.
References
- Moubray, J. Reliability Centered Maintenance, 2nd ed.; Industrial Press, Inc.: New York, NY, 1997.
- Wireman, T. Total Productive Management; Industrial Press, Inc.: New York, NY, 2004.
- Doberstein, T. Managed maintenance: more than just savings. Man. Mod. Lab. 2005, 7(4), 41–3.
Dr. Collins is Laboratory Services Manager,
Asset Management Services, Thermo
Electron Corp., 1410 Gillingham, Sugar
Land, TX 77478, U.S.A.; tel.: 713-272-2282; fax: 713-272-5334; e-mail: [email protected].