Continuous Improvement in Maintenance Strategies

06.5.2013
Industries are increasingly paying attention to maintenance efficiency optimizing the level
in reliability and availability of assets. Many of the improvements could be obtained using
new technologies and strategies to maximize service level and to reduce the maintenance
costs, as long as it is possible to identify the business areas where a leap in technology
could render and optimize the maintenance processes.

Aitor Arnaiz
IK4-Tekniker
Mantenimiento y Fiabilidad
aitor.arnaiz@tekniker.es

Egoitz Conde
IK4-Tekniker
Mantenimiento y Fiabilidad
egoitz.conde@tekniker.es

Aitor San Sebastián

Orona Elevator Innovation
Centre
asansebastian@orona-group.com

Xabier Sagartzazu
IK4-Ikerlan
Ingeniería Mecánica
xabier.xagartzazu@ikerlan.es

One of Europe’s large elevator companies, ORONA, is an example on this trend. In their
search for service excellence, starting from an optimal level in reliability and availability,
ORONA, together with IK4, has worked in the development of a continuous improvement
cycle, which is able to consider the quality of service as well as the maintenance efficiency
in a coherent way.
The decission process, which has been in use in ORONA since 2009, is using common
techniques and standards, together with a specific module developed for evaluation of
different strategies, keeping always in mind the possibility to implement new
technologies related to predictive maintenance.
Introduction
Performance improvements in the maintenance and conservation activities of assets are
measured by availability and operational reliability. They should be obtained preserving
maximum quality and safety levels and minimizing the costs.
In the current scenario of competitiveness, improvement efforts are essential to reach
high levels of effectiveness and efficiency in every company’s production or operational
department. The purpose is to achieve competitive advantage (in products or offered
services) based on different hard-to-copy aspects, i.e. know-how.
To obtain maximum performance, the organizations must be prepared for changes and
there are three interconnected areas in the change concept [1]:
 Processes, work fluxes to achieve the improvements (e.g. doing more preventive
work instead of corrective work, etc.).
 Technologies to facilitate or enable some processes.
 Organization and people within the organization must validate any change, so
there is a need of tools to ease changes.

Figure 1. Predictive actuation area for a normal probability distribution failure.

One of the approaches for improvement is to identify and to apply techniques and tactics
using high-tech elements, which would help to identify anomalies with high reliability.
Applying advanced technologies for the improvement of operation and maintenance
processes are usually nuanced or cancelled, partly because of a lack of tools for coherent
evaluation of the effects to costs and profits of the application (how it affects the life
cycle of the product or global productivity of the plant).
In this context, predicitive actuation is a current area of improvement. As show in FIGURE
1, habitual CBM systems are centred on the left area of distribution in order to mitigate
critical systems failures before to the time based replacement (or repair). It is important
to take into account the right area where more and more cost-effective predictive
systems are able to extend the repair periods.
Maintenance Strategy
In the search of excellence in service, ORONA has launched a research programme
focusing in three strongly related aspects:
 Product: Data acquisition with sensors connected to the control system, focusing
especially on the failure modes of critical components.
 Process: Definition of the maintenance processes using a better formalization of
these processes and personalized to each user.
 System: Allowing data exploitation to obtain information (historic data, control
data, etc.) by lift monitoring.
In order to deal with productprocess aspects, a project called “product and strategies
analysis of elevator conservation” was launched to establish improvement steps of
products related to maintenance processes. The main goal of the project is to show a
continuous search model of better processes using cost-effective technologies,
particularly emphasizing predictive maintenance strategies. New sensors,
communication systems, standards and protocols, concepts, etc. are continuously
emerging to the market.
Moreover, a permanent cost reduction of mature technologies enables their large-scale
implementation, following trend of improvement in the optimization of assets and
maintenance processes.
In recent years, there has been a substantial improvement in predictive technologies for
mechanical, thermal, electro- mechanical and more recently in electric and electronic
system monitoring [2].
Improvement Model
Maintenance should be a constantly improving activity, which enhances the quality of
service and optimizes operating costs. Condition-based maintenance and predictive
strategies based on cutting-edge technologies are arriving on the market and their
continuous cost reduction opens wide opportunities, helping the operation and
maintenance personnel to perform tasks more effectively.
There are cost-effectiveness studies of the different types of strategies [3,4,5,6], but it is
difficult to measure the impact of predictive strategies.
A simple model was developed to solve the difficulty of showing how predictive
maintenance could help in a cost-effective way. This model is based on the application of
different existing tools (Balance Scorecard, RAMS optimization, FMECA, FRACAS,…), and
it follows a six-step structure based on a Deming cycle to gradually improve each process
or service adapted to maintenance needs.
The cycle is continuously fed with new information to achievean optimal maintenance
working way in a cost-effective manner and keeping in mind strategies that are based on
predictive technologies. The steps are carried out in a cyclical manner as shown in figure
2.
1 – Selection of the objectives
The first step is to establish the main objective. It is essential to know exactly the situation
of the company in order to know what should be improved and to align the vision-
mission-strategies- objectives-indicators. These objectives should be identified with KPIs
(Key Performance Indicators) with different approaches: financial, learning or technical.
There are many different techniques that can serve to achieve the correct alignment
between the company and indicators, such as Balanced Scorecard.
2 – Identification of the most important products/processes
The next step is the identification of the main objects or processes where the
improvements are going to be critical with respect to their impact on the selected KPIs.
Results may include machinery parts (e.g. door mechanism), product types (e.g. semi-
automatic doors) or target sectors (e.g. hotels) among others. Criticality tables are used
in order to rank the results and select an appropriate subset for further analysis and
simulation in the next steps.
3 – Analysis of selected product/processes
An exhaustive analysis of selected products/processes is carried out to have a clear idea
of their main important aspects.
Analyzing the most critical assets is very useful in order toobtain the selected objectives.
The information of different tools: Failure Mode and Effect Analysis (FMEA), Risk Analysis
(PHA, HAZOP), Failure Tree Analysis (FTA) or Failure Reporting and Corrective Action
System (FRACAS) analysis can be used depending on the objectives and information
available. In conclusion, a complete understanding of these assets gives a better way to
make improvements.
4 – Development of the proper strategy for each critical product/ processes
This step consists of an analysis and an assessment of various maintenance strategies for
the selected critical products/ processes. There are a number of different techniques for
implementing and analysing these aspects.
The cost assessment simulation is done in this step and is the most critical part of the
whole model. An ad-hoc optimization tool has been developed to better simulate the
cost-effectiveness, and therefore this step will be further explained in this article.
5 – Implementation of actions
The selected strategy is implemented at least in one control group in order to evaluate
the results. New procedures, hardware and software technologies will be deployed and
tested.
6 – Final assessment
Using the initially defined indicators, the assessment will evaluate whether the initial
objectives have been fulfilled or not. Steps 5 & 6 are run iteratively in order to compare
the simulation results in step 4 with the real results obtained after the deployment of the
new strategies. This deployment is normally progressive (concept complete design, lab
trials, first installs on selected machines, etc.)
If the objectives are not fulfilled, i.e. there are large deviations from the simulated cost
assessment, it is necessary to return back to the previous step and identify the deviations
sources. If the objectives are being fulfilled, new objectives should be defined to follow
the continuous improvement programme.
Cost-Effectiveness Analysis
he cost-effectiveness analysis step is very important in this model, because it is the way
to indicate if any profit or competitive advantage can be achieved by using more
automatic maintenance tasks, especially predictive maintenance.
With this in mind, the analysis is done with a maintenance strategies simulator [8], which
introduces the reliability aspects of predictive systems with other common data as
corrective and preventive maintenance costs and reliability information. These aspects
appeared in [6] and recently have been studied by other authors [10,11].
The key issues in simulation are to use a valid source of information, employ a relevant
selection of key characteristics and behaviours, make approximations and assumptions
when necessary and understand the fidelity and validity of the simulation outcomes.
For the simulation of maintenance strategies, it is assumed that the probability density
distribution of failure for the equipment or component is known. This function describes
the possibility that a failure occurs at an instant of time. This is typically established from
data collected in trials, testing the items and noting the time when the failure occurs or
with reliability data, for example supplied by the manufacturer. The confidence value of
the simulation data is very important.

Figure 3. Multi-cycle optimization based in the scheme of figure 2.

Figure 4. ORONA’s door mechanism.

The improvement Cycle of Information Quality

Improvement cycle is not only focused on optimizing the maintenance processes, it is
also related to the process of information acquisition and storage as well as in the
identification of best indicators for finding deficiencies, which in other ways could be
difficult to analyse.
Outcomes of the Improvement Process
Among other improvements, the analysis allowed to focus on different aspects of the
cabin and floor doors considered most important at different cycles, such as certain door
mechanism (FIGURE 4). Selecting these critical elements allowed a detailed study of the
principal failures and causes, using information of FMEA combined with information
analysis of the lift population.
The analysis was focused on technologies with potential to identify the failure before it
happened and with the help of the work team knowledge different systems were found
to solve failures.
The improvement analysis finishes with the estimation of the potential impact in the
indicators with the current conditions of operation and maintenance. The impact of the
different strategies (corrective, timebased, inspections) is measured in Euro’s comparing
the maximum cost of the selected technologies and alternative strategies, considering
several known variables such as frequency of the inspections, reliability, cost of failures,
inspections and preventive actions.
In Figure 5 the y-axis describes the costs in Euro/month during the lift’s lifecycle, the
online monitoring costs (installation, maintenance) are simulated in the x-axis to check
the optimum cost for this type of CBM strategy. It is possible to check the maximum cost
of the online monitoring system by comparing it with the other maintenance options. In
this case, the graphic indicates different strategies and their cost-effectiveness.
Comparison between different strategies including the partial costs of the different
activities, are shown in figures 6 and 7.
Figure 5. Impact of the different strategies.

Figure 6. Inspection strategy with different frequencies shows how decreasing the
inspection frequency is not the right approach because it could produce savings, but there
will be a cost in quality of service (failures).
Outcomes of the information Improvement
The outcomes of the project are not only related to the identification of ‘hot spots’ where
to apply new technologies and strategies. The continuous improvement had also an
important impact in the quality of theinformation that is handled by ORONA services.
A new optimized visit report has been designed, both in manual and electronic versions.
The main goal is to reduce the amount of incidences not yet linked with a clear type of
failure. Vaguely defined fields of the work order were erased (“Other” type of failures)
and now there are three codes to be filled in: failure code, cause code and made action
code. Codes with little impact in the indicators were eliminated to have a manageable
work order for the maintenance staff.
First available results showed a great improvement of the identification of failures with
the three codes: failure/cause/action, there is now a clear identification of the failures
which is essential for suggesting improvements and speeding up work order fulfilment
(less notes are needed in the work order).
During the first steps of the project there was only one indicator related to the reliability
of the product. During the project development new indicators were discovered and
selected as the need to consider other lift costs and quality of service related aspects
became evident. Currently there are four indicators to identify the impact of the potential
improvements in maintenance.
The need of better information quality enabled identifying improvement points for
infrastructures associated to information access, transmission and processing. These
improvements are being handled by different groups in the company.

Impact in the Organization

The potential impact of the technologies is often reduced due to organizational issues:
 Even if the company has a dynamic organization, there is a rejection of new
technologies by maintainers, because it is seen as a risk to their job.
 New processes are difficult to fit into the organizational structure
In general it is very important to involve people from all levels of an organisation when
an improvement is implemented. Orona Elevator Innovation Centre’s e-conservation
team is working to integrate people from all areas of the maintenance business to adjust
the processes related to maintenance technologies.
Conclusions
The described process is expected to be a useful tool in different sectors (energy,
transport, production, SME…), where it is difficult to show the obtainable improvements
when implementing new processes or products, particularly in predictive maintenance
(e.g. wear control, inspection system improvement, etc.).
Although predictive strategies are shown to be cost-effective for companies, the lack of
understanding these techniques, initial costs and reticence for maintenance culture
change, require tools to prove the cost-effectiveness.
In conclusion, this type of research is relevant in the area of continuous technological
improvement, where today’s experimental technology could be tomorrow’s competitive
advantage ensuring the optimal level in reliability and availability with the lowest possible
cost.
Figure 7. Smart combination of on-line monitoring and on-site inspections (right) ensures
maximum quality of service (less failures) while notably reducing overall costs.
»»References ›› [1] Bengtsson M & Salonen A. “On the Need for Research on Holistic
Maintenance”, Figure 7. Smart combination of on-line monitoring and on-site inspections
(right) ensures maximum quality of service (less failures) while notably reducing overall
costs. Proceedings of 22nd International Congress on Condition Monitoring and
Diagnostic Engineering Management 2009, pp. 165–172. ›› [2] Vichare N & Pecht M.
“Prognostics and Health Management of Electronics”, IEEE Transactions on Components
and Packaging Technologies, Vol. 29, No. 1, March 2006, pp. 222–229 ›› [3] B. Al-Najjar.
Economic Criteria To Select A Costeffective Maintenance Policy, Journal of Quality in
Maintenance Engineering, Vol. 5 Number 3, 1999 ›› [4] H. Fromm, S. Heck & M. Buderath.
Cost Benefit Analysis Of A Health Management System, 1st CEAS European Air and Space
Conference, 2007. ›› [5] Feldman K, Jazouli, T & Sandborn P. (2009) A Methodology for
Determining the Return on Investment Associated with Prognostics and Health
Management. IEEE Transactions on Reliability, Vol. 58, Nº 2, pp. 305–316. ›› [6] Jardine,
A.K.S. & Tsang, A.H.C. “Maintenance, replacement and reliability: Theory and
Applications” CRC Press 2005 ›› [7] Parida A. “Study and analysis of maintenance
performance indicators (MPIs) for LK - AB” Journal of Quality in Maintenance Engineering
Vol. 13 No. 4, 2007 pp. 325–337 ›› [8] Conde E, Fernandez S, Arnaiz A., Maintenance
methodology for selecting the most cost-effective predictive strategies. The 37th ESReDA
Seminar, Octubre 2009 Baden (Suiza). ›› [9] Jardine, A.K.S.; Joseph, T.; Banjevic, D.
“Optimizing condition-based maintenance decisions for equipment subject to vibration
monitoring”, Journal of Quality in Maintenance Engineering, Vol. 5 No. 3, pp. 192–202,
1999 ›› [10] Al-Najjar B. (2009) A computerised model for assessing the return on
investment in maintenance; Following up maintenance contribution in company profit.
Proceedings of the Fourth World Congress on Engineering Asset Management (WCEAM)
2009, pp. 137–145, Greece, Athens, Sep. 2009-09-08, Springer-Verlag London Limited. ››
[11] Berdinyazov A, Camcj F, Sevkli M and Baskan S, “Economic Analysis of Maintenance
Policies for a System”. IEEE International Symposium on Diagnostics for Electric
Machines, Power Electronics and Drives, 2009. SDEMPED 2009.