Proceedings of the IISE Annual Conference & Expo 2022
K. Ellis, W. Ferrell, J. Knapp, eds.
OEE Data Visualization and Analysis – An Industry 4.0 Perspective
Abstract ID: 1145947
Ahmad E. Elhabashy and Hadi Fors
Production Engineering Department, Alexandria University, Alexandria, Egypt
Abstract
The Overall Equipment Effectiveness (OEE) metric has been long used for determining the productivity of machines
in manufacturing systems. Centered on Total Productive Maintenance principles, the OEE metric is the product of a
machine’s availability, performance, and quality. Each of these three factors are typically calculated from different
sets of information. With the abundance of data obtained from the production floor nowadays resulting from the
ongoing technological advancements, important information related to the OEE metric and its factors could easily be
overlooked. To this end, this work proposes an Industry 4.0 perspective for presenting OEE data, allowing better
visualization and analysis. As an implementation, data from a single machine within a short period of time in a
practical industrial setting was used. Based on the performed analysis, recommended actions were provided to improve
production efficiency and reduce quality issues.
Keywords
Big Data, Data Analysis, Data Visualization, Industry 4.0, Overall Equipment Efficiency (OEE).
1 Background and Motivation
In today’s everchanging industrial environment, manufacturing firms need to adapt to the various changes in the
surrounding environment to remain competitive enough. Such changes are often attributed to shifts in customer
preferences, government regulations, technology advancements, and increasing competition. Furthermore,
manufacturing firms typically seek to maintain a high level of profitability and a large market share through achieving
goals such as reducing costs, increasing productivity, efficient deployment of resources, and decreasing losses.
Of these different goals, increasing productivity and decreasing production losses is the focus of this work as they
would yield significant time and money savings. There typically exists large amounts of losses in a manufacturing
system, affecting both the machine’s efficiency and the product’s quality. These losses could be due to manufacturing
disturbances that are either chronic (small and hidden) or sporadic (more obvious and rapid) [1]; and are broadly
grouped into either downtime, speed, or quality losses [2]. One approach that would help with the goal of decreasing
production losses is applying Total Productive Maintenance principles, which involve preventive maintenance
practices, total quality control concepts, and employee involvement [3].
Total Productive Maintenance (TPM) is usually applied to increase equipment’s productivity, reduce maintenance
costs, and reduce machine breakdowns [4]. Centered on TPM concepts, the Overall Equipment Effectiveness (OEE)
metric has been developed and long used for determining the productivity of an equipment in manufacturing systems
[5]. The OEE metric is the product of a machine’s availability, performance, and quality [5]–[9], with each of these
three factors corresponding to a production loss group and typically calculated from different sets of data.
With the emergence of the fourth industrial revolution and its related technologies, the method of obtaining OEE data
from the production floor and its size is constantly changing. The fourth industrial revolution is often referred to as
“Industry 4.0”, after the initiative proposed in 2011 at the Hannover Fair in Germany [10], [11]. Industry 4.0 aims to
achieve better operational efficiency, productivity, and automation [10] and it encompasses different technologies,
such as the Industrial Internet of Things (IIoT), Cloud computing, Big Data Analytics, and Blockchain [10]–[15]. As
an example, Big Data Analytics refer to the advanced technologies that are able to deal with a considerable amount of
data, which is too large to handle using traditional techniques, to get any value out of it [11].
Elhabashy and Fors
In the presence of such technologies, OEE data no longer need to be collected manually; but could be acquired from
various sensors at an almost continuous rate. Nowadays, data for OEE calculations is readily available in large amounts
[6]. On one hand, this data could be used for online monitoring in order to optimize production, suggest corrective
interventions, and provide preventive maintenance alerts. On the other hand, with such abundancy of available data,
important information related to the OEE metric and its factors could easily be overlooked.
Improving manufacturing systems’ efficiency and reducing the production losses, indicated by the OEE metric
enhancement, is the overarching goal of this research work. More specifically, the aim of the effort presented in this
paper is to offer an Industry 4.0 perspective for presenting OEE data in a form allowing better visualization and
analysis in practical situations, which would be suitable when dealing with Big Data. Based on the performed analysis,
recommended actions would be provided to improve production efficiency and reduce quality issues.
This paper is organized as follows; in section 2, the approach followed for visualizing and analyzing OEE metric is
presented. Section 3 includes the implementation of the approach on a single machine in a tire manufacturing
company, and finally the paper is concluded and future work ideas are illustrated in the last section.
2 Approach
The approach followed in this work is based on incorporating OEE calculations into common Big Data Analytics
procedures. The steps of this approach are illustrated in Figure 1 and briefly discussed in the following sub-sections.
Requirements Data Data Data
Data Analysis
Specification Collection Processing Visualization
Figure 1: Suggested approach steps for visualizing and analyzing OEE metric.
2.1 Requirements Specification
The first step is to identify the necessary data for the analysis, which is the input data for the OEE metric calculations
in this work. As previously mentioned, the OEE data is the product of the three factors shown in equation (1).
𝑂𝐸𝐸 = 𝐴𝑣𝑎𝑖𝑙𝑏𝑖𝑙𝑖𝑡𝑦 × 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 × 𝑄𝑢𝑎𝑙𝑖𝑡𝑦 (1)
The first factor, the availability, considers all the events that stopped the production process and is the ratio of the
available time to the loading time. The latter is just the planned production time, whereas the former is the loading
time without the downtime [1], [6], [8]. The downtime results from either unplanned stops (e.g., breakdowns) or
planned stops (e.g., changeovers) [16].
𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑙𝑒 𝑇𝑖𝑚𝑒 𝐿𝑜𝑎𝑑𝑖𝑛𝑔 𝑇𝑖𝑚𝑒 − 𝐷𝑜𝑤𝑛𝑡𝑖𝑚𝑒
𝐴𝑣𝑎𝑖𝑙𝑏𝑖𝑙𝑖𝑡𝑦 = = (2)
𝑃𝑙𝑎𝑛𝑛𝑒𝑑 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑇𝑖𝑚𝑒 𝐿𝑜𝑎𝑑𝑖𝑛𝑔 𝑇𝑖𝑚𝑒
The second factor, which is the performance, takes into account the conditions causing the manufacturing process to
run less than its normal speed and is the ratio of net operating time to the available time [1], [5], [6], [8]. The net
operating time is just the ideal cycle time multiplied by the total number of manufactured products, where the ideal
cycle time can be defined as the minimum cycle time that a process is expected to achieve in ideal circumstances [16].
𝑁𝑒𝑡 𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑛𝑔 𝑇𝑖𝑚𝑒 𝐼𝑑𝑒𝑎𝑙 𝐶𝑦𝑐𝑙𝑒 𝑇𝑖𝑚𝑒 × 𝑇𝑜𝑡𝑎𝑙 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑠
𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 = = (3)
𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑙𝑒 𝑇𝑖𝑚𝑒 𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑙𝑒 𝑇𝑖𝑚𝑒
Quality, which is the final factor, considers the manufactured products that do not meet quality standards and is just
the ratio between non-defective products and the total products [1], [5], [6], [8].
𝑁𝑜𝑛 − 𝑑𝑒𝑓𝑒𝑐𝑡𝑖𝑣𝑒 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑠 𝑇𝑜𝑡𝑎𝑙 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑠 − 𝐷𝑒𝑓𝑒𝑐𝑡 𝐴𝑚𝑜𝑢𝑛𝑡
𝑄𝑢𝑎𝑙𝑖𝑡𝑦 = = (4)
𝑇𝑜𝑡𝑎𝑙 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑠 𝑇𝑜𝑡𝑎𝑙 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑠
Elhabashy and Fors
2.2 Data Collection
In this step, the required data is gathered, either manually or automatically. Data is usually collected from various
sources, such as databases, observations from the production floor, sensors, etc. The data obtained may not be
structured and could contain noise or irrelevant information, requiring processing before being further used for any
application.
2.3 Data Processing
The collected data would then need to be processed, such as requiring organizing and filtering. This step may include
structuring data into a certain format for the subsequent steps, such as converting from rows to columns or vices vera.
In addition, the obtained data may be incomplete or containing duplicates and/or errors, requiring filtering. In this
work, MS Excel is used for data processing, as it allows quick calculations and contains a wide range of built-in
functions that are extremely useful in this step.
2.4 Data Visualization
After data is organized and filtered, the data is visualized through graphs and other displays. Data visualization help
practitioners with monitoring manufacturing equipment, better understanding how the production system works, easily
assessing its performance, and quickly taking necessary interventions if needed.
For this step, MS Excel was not used as it is not quite effective in visualizing Big Data and producing customized and
interactive displays using MS Excel is challenging. Instead, Microsoft Power BI software was used for data
visualization in this work. Power BI has several advantages, such as 1) providing better visuals, 2) alerting users when
specific metric values change, 3) being a great option for dashboards and Key Performance Indicators (KPIs)
representation, and 4) allowing dashboard interactivity, via different types of filters. It should be noted that Power BI
was not well suited for data processing and, hence, MS Excel was used in that step.
2.5 Data Analysis
The purpose of this final step is to gain useful insights from the visualization of the collected data, enabling decision-
makers to better understand, interpret, and derive conclusions about the production system and propose corrective
actions. For example, data analysis could highlight losses that delay the production process, recurring issues, and
ongoing trends. The analysis was performed using the Power BI software, as it also contains data analysis capabilities.
3 Implementation
3.1 Company Overview and Current Situation
The approach discussed in the previous section has been applied to track, visualize, and analyze OEE of a single
machine in a tire manufacturing company. The company manufactures branded truck, bus, agricultural vehicles, and
off-road tires. The company has two factories; the first factory includes mixer and semi-finishing departments, while
the second factory includes building, curing, and inspection departments. The studied machine is fabricating the bead
of the tire in the semi-finishing department. This machine is used in manufacturing three different sizes of tires.
The company had provided the OEE data, but it was in a specific form that doesn’t allow it to be easily visualized and
analyzed. The necessary data for the analysis was then identified (e.g., number of shifts, shift duration, planned stops,
etc.). Also, the required inputs for analysis related to the studied machine were determined (e.g., ideal cycle time for
different products, number of manufactured products, amount of scrap, unplanned stops, etc.).
3.2 Data Processing and Visualization
The collected data were processed since there was incomplete data, it contained duplicates and errors, and was difficult
to analyze directly. The collected data were then reorganized, filtered, and stacked in columns by using MS Excel.
After that, data visualization was conducted using Power BI to obtain insights from the information within the data.
Two dashboards, a primary and secondary one as displayed in Figure 2 and Figure 3, were designed to show visuals
related to OEE and its factors such as availability, performance, and quality. The description of main and secondary
dashboard sections is presented in Table 1. It should be noted that the dashboards include both date and shift filters
to allow for more insights about the machine data.
Elhabashy and Fors
Figure 2: Dashboard Main Display
Figure 3: Dashboard Secondary Display
Table 1: Main & Secondary Dashboard Sections’ Description
# Description # Description
1 OEE Average Date Filter (Filter Dashboard by A
9, 15
2 Availability Average Specific Date/Period)
3 Performance Average Shift Filter (Filter Dashboard by A
10, 16
4 Quality Average Specific Shift(s))
5 Daily OEE and its three factors 11 Total Unplanned Stops of Different Types
6 Total Count of Products 12 Scrap Percentage of Each Product
7 Key Influencers Affecting OEE 13 Daily Unplanned Stops
Total Planned Stops, Unplanned Stops, 14 Daily Total Amount of Scrap
8 Non-Declared Time, Total Time, Loading
Time, and Available time
Elhabashy and Fors
In addition, the following visuals were developed to allow for more insights about the studied machine.
• Key influencer visual: determines the influence of changing the three factors on the OEE metric value and
which factor has greater influence on the OEE metric.
• Time visuals: show the different times related to the machine (total time, loading time, available time,
planned stops, unplanned stops, and non-declared time).
• Total count visual: shows the total number of manufactured products of each type.
• Losses analysis visuals: could be used to know the most occurring planned and unplanned stops and when
they occur over the month. The visuals also show the machine stops related to availability, performance, or
quality separately.
• Scrap analysis visuals: are helpful in illustrating the change in amount of scarp over the month and the scrap
percentage within the manufactured products for all types.
Due to the COVID-19 circumstances, companies nowadays are searching for online alternatives to enable managers
and decision makers to work remotely. A mobile application and website were developed to be suitable online
alternatives. The dashboards can be retrieved from a mobile application which facilitates remote monitoring. Also,
the dashboard or a part of the dashboard can be accessed online by authorized users through a website without the
need to have the application.
3.3 Data Analysis and Recommended Actions
The visualized data was analyzed to show the effect of changing the three factors on the OEE metric and suggest
recommended actions to improve it. In general, the OEE metric could be enhanced by increasing the available time
which can be achieved by reducing the planned or unplanned stops. Also, decreasing the number of scrapped products
helps in increasing OEE as well. The following points summarizes the analysis and recommended actions related to
OEE factors, planned stops, unplanned stops, and scrap.
• Key OEE Factors: As shown in the main dashboard (section number 7), the first factor that needs to be
addressed is performance, as it has higher impact on increasing the OEE. Also, the dashboard revealed that
increasing the performance by 42% leads to improving the OEE by approximately 25%, increasing the
availability by 34% leads to improving the OEE by approximately 7%, and increasing the quality by 41%
leads to improving the OEE by approximately 3%. It should be noted that it may not be practically possible
to achieve a quality factor of more than 100%, as suggested.
• Planned Stops: As illustrated in the main dashboard (section number 8), the total time of planned stops over
one month was 8,845 minutes which represents 19.8% of the total time. It was noticed that the main planned
stops were lunch rests, coffee breaks, non-programmed (time for adjusting the machine at the beginning of
the shift in case of producing different products), and planned absents. The first two losses cannot be reduced.
For the non-programmed losses, they can be reduced by training the employees to adjust the machines in less
time. For the planned absents, appropriate labor scheduling is needed to avoid having planned absents for the
same positions on the same day.
• Unplanned Stops: As presented in the main dashboard (section number 8), the total time of unplanned stops
over one month was 6,169 minutes which represents 13.8% of the total time. It was observed that the top
four unplanned stops were waiting for material, breakdowns, start/end work, and setup. Regarding waiting
for material time loss, it is recommended to perform a facility planning study to optimize the space between
the machine and material inventory to reduce transportation time or provide the required material for the
machine at the right time to decrease this loss. Breakdowns could be reduced by applying preventive
maintenance based on historical data of the machine breakdowns or perform a quick inspection on the
machine at the beginning of each shift. Start/end work losses could be minimized by giving incentives to
labors based on number of products they produce per shift. Setup time could be reduced by applying SMED
(Single Minute Exchange of Dies) concept.
• Scrap: In the secondary dashboard (section number 12), the scrap percent of each product is presented. It
was observed that the last product has the highest percent (6%). Accordingly, it is recommended to perform
a root cause analysis to reveal the main reasons for producing the defective products.
4 Summary and Future Work
In this work, an Industry 4.0 perspective for displaying OEE data, allowing improved visualization and analysis was
presented. As an implementation of the proposed approach, large amounts of data (Big Data) from a single machine
within a short period of time (one month) in a tire company was used. The collected data were filtered, organized, and
Elhabashy and Fors
processed using Excel. Then, the data was visualized through a dashboard with two displays using Power BI. After
that, the data visualization via the software allowed analyzing the main factors affecting the OEE metric. Based on
this analysis, recommended actions to help in reducing losses and improving the production process were suggested.
As for the future of this research work, there are two potential tracks: 1) improving the data visualization and 2)
enhancing the data collection process. Regarding the first track, improvements could involve expanding the scope of
the dashboard to include data from a complete production department or even the whole company, designing the
dashboard to give real-time alerts when specific variables exceed critical pre-specified values, and incorporating
artificial intelligence techniques into the dashboard to have the capability to suggest appropriate corrective actions.
One key idea for future work within the second track in this specific company is taking advantage of the IIoT concept
by adding data collection sensors (or cameras) on the machines to monitor important machine variables, such as feeds,
speeds, wear in mechanical components, and shortage of material. This would improve the quality of the collected
data and the production process by providing real-time indication for the process variables.
Acknowledgements
The authors would like to thank the members of the 2020-2021 senior design project team for their role in this work:
Abdelrahman Fadel, Abdelrahman Fathy, Ahmed M. Abdelhamed, Alaa Fayed, Sarah Elsayed, and Somaya Yasser.
References
[1] P. Jonsson and M. Lesshammar, “Evaluation and improvement of manufacturing performance measurement
systems ‐ the role of OEE,” Int. J. Oper. Prod. Manag., vol. 19, no. 1, pp. 55–78, Jan. 1999, doi:
10.1108/01443579910244223.
[2] L. Pintelon, L. Gelders, and F. Van Puyvelde, “Maintenance Management,” Omega-International Journal Of
Management Science, vol. 18, no. 1. Acco, 1997, doi: 10.1016/0305-0483(90)90018-5.
[3] S. Nakajima, “Introduction to TPM: total productive maintenance.(Translation),” Product. Press. Inc., 1988,
p. 129, 1988.
[4] A. Jain, R. Bhatti, and H. Singh, “Total Productive Maintenance (TPM) Implementation Practice,” Int. J. Lean
Six Sigma, vol. 5, no. 3, pp. 293–323, Jan. 2014, doi: 10.1108/IJLSS-06-2013-0032.
[5] P. Muchiri and L. Pintelon, “Performance Measurement using Overall Equipment Effectiveness (OEE):
Literature Review and Practical Application Discussion,” Int. J. Prod. Res., vol. 46, no. 13, pp. 3517–3535,
Jul. 2008, doi: 10.1080/00207540601142645.
[6] P. Ghafoorpoor Yazdi, A. Azizi, and M. Hashemipour, “An Empirical Investigation of the Relationship
Between Overall Equipment Efficiency (OEE) and Manufacturing Sustainability in Industry 4.0 with Time
Study Approach,” Sustainability, vol. 10, no. 9, p. 3031, 2018.
[7] P. N. Muchiri, L. Pintelon, H. Martin, and A.-M. De Meyer, “Empirical Analysis of Maintenance Performance
Measurement in Belgian Industries,” Int. J. Prod. Res., vol. 48, no. 20, pp. 5905–5924, Oct. 2010, doi:
10.1080/00207540903160766.
[8] A. J. De Ron and J. E. Rooda, “OEE and equipment effectiveness: an evaluation,” Int. J. Prod. Res., vol. 44,
no. 23, pp. 4987–5003, Dec. 2006, doi: 10.1080/00207540600573402.
[9] S. J. Benjamin, M. S. Marathamuthu, and U. Murugaiah, “The use of 5-WHYs technique to eliminate OEE’s
speed loss in a manufacturing firm,” J. Qual. Maint. Eng., vol. 21, no. 4, pp. 419–435, Jan. 2015, doi:
10.1108/JQME-09-2013-0062.
[10] Y. Lu, “Industry 4.0: A survey on technologies, applications and open research issues,” J. Ind. Inf. Integr.,
vol. 6, pp. 1–10, 2017, doi: https://doi.org/10.1016/j.jii.2017.04.005.
[11] M. Ghobakhloo, “The future of manufacturing industry: a strategic roadmap toward Industry 4.0,” J. Manuf.
Technol. Manag., vol. 29, no. 6, pp. 910–936, Jul. 2018, doi: 10.1108/JMTM-02-2018-0057.
[12] R. Y. Zhong, X. Xu, E. Klotz, and S. T. Newman, “Intelligent Manufacturing in the Context of Industry 4.0:
A Review,” Engineering, vol. 3, no. 5, pp. 616–630, 2017, doi: https://doi.org/10.1016/J.ENG.2017.05.015.
[13] S. Vaidya, P. Ambad, and S. Bhosle, “Industry 4.0 – A Glimpse,” Procedia Manuf., vol. 20, pp. 233–238,
2018, doi: https://doi.org/10.1016/j.promfg.2018.02.034.
[14] P. K. Muhuri, A. K. Shukla, and A. Abraham, “Industry 4.0: A bibliometric analysis and detailed overview,”
Eng. Appl. Artif. Intell., vol. 78, pp. 218–235, 2019, doi: https://doi.org/10.1016/j.engappai.2018.11.007.
[15] E. Oztemel and S. Gursev, “Literature review of Industry 4.0 and related technologies,” J. Intell. Manuf., vol.
31, no. 1, pp. 127–182, 2020, doi: 10.1007/s10845-018-1433-8.
[16] Vorne, “Calculate OEE.” https://www.oee.com/calculating-oee/.
Reproduced with permission of copyright owner. Further reproduction
prohibited without permission.
