How to Spot and Reduce Bottlenecks in Production and Packaging EOL?

Achieving the excellence in packaging process and operation is the ultimate goal of implementing packaging equipment and EOL (End-of-Line). However, production would not always work in a perfect status at the fastest speed, producing only good parts and without any unexpected downtimes [PackIoT, December 20, 2019, 1] – that the whole packaging end-of-line reaches an ideal operational status of 100% OEE (Overall Equipment Effectiveness).

While having the dream come true is what everyone expects, the harsh reality is that – this is hardly possible a situation in any packaging end-of-line.

In this article, we dig into the underground facts about:

• What are the bottlenecks in the production packaging circumstances?
• How to find out and spot the bottlenecks in production?
• How to avoid or improve bottlenecks?

What Does Bottleneck Mean in Production?

The literal definition of bottleneck refers to the narrow route of a path that causes a point of congestion.

In the scene of production like packaging automation, processing, and manufacturing, bottlenecks refer to the stages or processes that limit the overall output of the production system where the flow of work or materials is impeded, causing delays or inefficiencies that affect the overall productivity and output.

Or say, they are points among manufacturing and production processes and systems where it occurs that the workload exceeds the capacity of a particular section, equipment, process, and procedure, or other devices. and sometimes as well as people. [Patrick Lemay 2022, 2]

The bottlenecks in packaging EOL are not such a simple issue compared with that of a manual process where such bottlenecks can be reduced by investing in hiring more employees. Rather, identifying the bottleneck in production and packaging EOL is a far more complex topic and requires certain professionalism to solve the problem. In some cases, the bottlenecks are not about the equipment but the operational processes.

What are the Causes of Bottlenecks in the Production and Packaging EOL?

Identifying production bottlenecks is a critical step of BA (Bottleneck-Analysis) which is one of the essential lean manufacturing concepts. [3]

When it comes to identifying bottlenecks with production, various factors can be the culprit including equipment that is incorporated along the packaging EOL, sometimes people who operate those machines, or even factors related to management and operational processes.

Causes of Production Bottlenecks
Credit: ELITER Packaging Machinery

Xingjian Lai at the University of Michigan once lists above potential causes of bottlenecks in one of his papers that studies the graph neural network approach for production bottlenecks. They can be grouped into 3 categories as follows:

• Causes of bottleneck due to machine
• Causes of bottleneck due to people
• Causes of bottleneck due to management and operational process

Failing to address the types of bottlenecks can result in your productivity constantly falling behind your expectations. To look into the details and to put some examples of each potential cause of the bottlenecks mentioned above:

How Machine’s Status May Cause Bottlenecks in Production?

Bottlenecks in production and packaging EOL caused by machines are primarily about how the machine has been maintained, their maximum capability, and how such a capability has been taken avail of.

Pick-and-Place Station of a Top-Load Case Packing System

Downstream machines may be a bottleneck of the EOL to limit its overall capacity and speed. For example, Danone – a multinational food and beverage company based in France, may sign with ELITER Packaging Machinery to work on a yogurt packaging end-of-line consisting of a sleever and then a top-load case packer. The project requires yogurt cups welded together in format 2×3 to be wrapped by carton sleeve and then processed for case packing in format 2×3 and 2 layers.

While our wrap-around sleever is capable of running at a speed of 135 packs per minute, an output after once divided by the number of yogurt packs in each corrugated cardboard box would be the required speed for the mentioned top-load case packer, which is for some 10 cases per minute.

A bottleneck here is about the mechanisms of the pick-and-place station of the top-load case packer. Given the speed of 10 cases per minute, the pick-and-place mechanism should then be able to reach 20 cycles per minute now that each corrugated box comes with 2 layers of yogurts inside – which exceed the speed limit of such a system.

As an inevitable result, the sleever installed at the upstream should be configured then with a slower speed to match the maximum capability of the top load case packing system.

How People May Cause Bottlenecks in Production?

Skilled and trained people are also an essential factor in reducing the bottlenecks in production.

The proficiency of workers with format and size changeover may be the cause of production bottleneck. Packaging EOL is frequently designed with options to cover different product sizes and formats for which it is required to do changeovers when the production is switching between different recipes. Regardless of the fact that some advanced packaging machines offer the option of motorized and automated changeover, most companies would not afford such flexible equipment but opt for manual intervention for size changeover. In such a case, the proficiency of the operator in terms of changeover counts significantly for the time consumed as part of the downtime of production.

Technicians in charge of the machine’s maintenance are critical factors to reduce bottlenecks in production. Technicians are responsible for regular maintenance activities such as cleaning, lubricating, and inspecting machines. This proactive approach helps identify and resolve minor issues before they escalate into major breakdowns, preventing unplanned downtime and bottlenecks in production.

How Process and Mangement May Cause Bottlenecks in Production?

Bottlenecks in production and management processes are rather more complex factors because they involve a broad range of reviews as well as impacts from 3rd parties including suppliers and even customers. Management bottlenecks can be further defined as either external bottlenecks or internal bottlenecks.

How to Identify Bottlenecks in Production and Packaging EOL?

BI (Bottleneck Identification) is the first step of BA (Bottleneck Analysis) to help remove or ameliorate production holdups and improve the overall effectiveness of the production. [4]

As one of the lean manufacturing philosophies, identifying bottlenecks and conducting improvement can substantially boost the productivity of a company yet this is by no means an easy task since bottlenecks in production, either internal or external, are often shifting and hard to figure out. Depending on the skills possessed by the leadership, the BI process can be conducted by several approaches including

• experimental approach,
• quantitative analysis,
• and by applying theoretical and methodological tools such as those provided by Six Sigma.

Quantitative and Formulated Analysis for Bottleneck Identification

Defining the problem and converting it into numbers for analysis is the critical mindset of modern management theory and methodologies provided by lean management and manufacturing.

A lot of packaging machines are nowadays incorporated with DSaaS and Industry 4.0 to collect data through operation and provide insight with the analysis tools that the user can review through the smart HMI.

Those who are familiar with that analytic tool may first conduct a process mapping to describe the structure of their serial production system to represent their production process, and then carry out data collection with regard to the appropriate KPIs (Key Performance Indicators) that are optimal to reveal the behavior of the system the detect the deviations. [5]

Process Mapping of a Production Line
Credit: ELITER Packaging Machinery

• Modeling and Formulation of 2-machine Case

The formulation and mathematical modeling of the above production sequence and mapping should start with defining the interactions between machines and buffers. A simplified study of this production sequence is a case of 2 machines with 1 buffer combined and modeled as a simple Markov Chain. Before such a study is conducted, the following concepts should be denoted from the modeling:

Markov Chain
Credit: Victor Powell

• Machines are denoted as \( (M_1, M_2, … , M_m) \) and buffers are referred to as \( (B_1, B_2, … , B_\text{M-1}) \)
• The cycle times of each machine along the production sequence are identical and the time frame in this modeling is confined to cycle time
• For each machine denoted in this sequence, each of them is subject to either Failure Status denoted as \( F \) or operating status denoted as \( U \)
• The probability that the machine keeps operating or that t is paused due to a fault after each cycle is denoted as \( P_k \)
• Machine can be fixed after a fault and back to operating at the beginning of the next cycler with probability as \(R_k\)
• All buffers here have a certain limited capacity and at the end of each cycle, the number of products within each buffer will change
• Machine will be blocked if it is operating while the buffer is fully taken by products at the beginning of a cycle time
• Machine will be operated but without products fed if at the beginning the each cycle time the buffer is empty – called the state of being starved

The 2-machine-1-buffer case can be molded with Markov Chain expressed as \( \Omega = \begin{Bmatrix} S1, n, S2 \end{Bmatrix} \) where \(S1 \) and \(S1\) stand for the status of the two machines – either operating or paused due to fault, while \( n \) stands for the number of products in the buffer between, denoted as \(B1\).

Define \( PR \) as the production rate of the machine and the probability at which \(M1\) is blocked as \(BL\), then starved machine \(M2\) at the probability of \(ST\), represented as follows:

• \(PR = \mathcal P \mathcal r \begin{vmatrix} \begin{Bmatrix} m_2 \text{  operating} \end{Bmatrix} \cap \begin{Bmatrix} B_1 \text{not empty} \end{Bmatrix} \end{vmatrix}, \)
• \(ST= \mathcal P \mathcal r \begin{vmatrix} \begin{Bmatrix} m_2 \text{  operating} \end{Bmatrix} \cap \begin{Bmatrix} B_1 \text{empty} \end{Bmatrix} \end{vmatrix}, \)
• \(BL = \mathcal P \mathcal r \begin{vmatrix} \begin{Bmatrix} m_1 \text{  operating} \end{Bmatrix} \cap \begin{Bmatrix} B_1 \text{ full} \end{Bmatrix} \cap \begin{Bmatrix} m_2 \text{ down} \end{Bmatrix} \end{vmatrix}, \)

which can be further defined as:

$$ \widehat {PR} = \frac {R_2}  {P_2 + R_2} \begin{bmatrix} 1 – \Phi (P_1, R_1, P_2, R_2, N)  \end{bmatrix} $$

$$ \widehat {ST} = \frac {R_2}  {P_2 + R_2} \cdot \Phi (P_1, R_1, P_2, R_2, N)   $$

$$ \widehat {BL} = \frac {R_1}  {P_1 + R_1} \cdot \Phi (P_2, R_2, P_1, R_1, N)   $$

where,

$$ \Phi(P_1, R_1, P_2, R_2, N) =
\begin{cases}
\frac {P_1 \theta_2} {(R_1 + R_2 – R_1 R_2)(P_1 + R_1)}& & if \,\,N=1 \\
& & \\
\frac {P_1 \eta_1 \eta_2 \theta^2_2 (P_2 + R_2)} {D_1 + D_2 + D_3 + D_4}&  & if \,\,N \gt 1 \\
\end{cases}
$$

The above formulations serve as a tool to measure the system and sequence’s performance and link status of a single machine, buffers and another to identiify the bottlenecks.

Identify Bottlenecks in Production with Theoretical Tools

Modern management theories like Lean Manufacturing and Six Sigma provide bunches of tools with which the leadership can figure out the bottlenecks in production.

• DMAIC

DMAIC is a compound analysis tool provided by Six Sigma that combines steps of (Define, Measure, Analyze, Improve, and Control) to eliminate waste, decrease variation in manufacturing, avoid defects, and streamline a business’s operational process – a roadmap to continuous improvement for businesses. DMAIC engages the transforming of a certain problem into a data-based question for the purpose of measurement, tracking results, and analysis.

DMAIC
Credit: ELITER Packaging Machinery

• Ishikawa Cause-Effective Diagram

The Ishikawa Cause-Effect Diagram, also known as the fishbone diagram or the Ishikawa diagram, is a visual tool used to identify and categorize potential causes of a problem or an effect. It was developed by Dr. Kaoru Ishikawa, a Japanese quality control expert, and is widely used in quality management and problem-solving methodologies such as Six Sigma.

• SIPOC Diagram

A SIPOC diagram is a visual tool used in process improvement and Six Sigma methodologies to understand and define the high-level process elements. SIPOC stands for Suppliers, Inputs, Processes, Outputs, and Customers.

By applying SIPOC diagram for bottleneck identification, the company is able to get an overview of either internal or external processes and to find out the culprit of the problem – either due to their own procedure or for the fault of their external stakeholders.

• OEE (Overall Equipment Effectiveness)

OEE is a metric commonly used in manufacturing to measure the efficiency and productivity of a piece of equipment or a production line. OEE takes into account three key factors: availability, performance, and quality. By analyzing the OEE data and identifying areas with low scores or consistent underperformance, manufacturers can pinpoint potential bottlenecks in production and implement targeted improvement measures.

How to Mitigate and Eliminate Bottlenecks in Production and Packaging EOL?

Following the bottleneck identification comes the step of measuring and elaborating actions for reducing and eliminating the detected bottlenecks.

This is more like a case-by-case problem since each production or manufacturing facility has its own site plan, implementation of EOL, and organizational structure and the root cause may vary and thus different strategies may be adopted. In general, the actions to eliminate bottlenecks may include:

• Streamline processes: Analyze and optimize the workflow within the bottleneck area. Look for any unnecessary or redundant steps that can be eliminated, reorganize workstations or equipment layouts to enhance efficiency, and standardize processes to reduce variability and errors.
• Train and empower employees: Provide proper training and empower your workforce to tackle bottlenecks. Equip employees with the skills and knowledge needed to identify and resolve issues in real time. Encourage open communication and collaboration among team members to foster a continuous improvement mindset.
• Implement lean manufacturing principles: Adopt lean manufacturing principles and techniques such as Just-in-Time (JIT) inventory management, 5S methodology, and value stream mapping. These methodologies can help identify and eliminate waste, improve flow, and optimize production processes, reducing bottlenecks.
• Adopt industry 4.0 technology to facilitate daily operation: Technology companies now offer a lot of options to incorporate machines and equipment with data tracking and collection systems as well as analysis tools with which leadership of manufacturing and production can keep tight control over their production sequence, process, and outputs.
• Working with a professional manufacturer to constantly offer improvement to the packaging machine installed: For a company like ELITER Packaging Machinery who offers packaging machines for the client, it is a long-term commitment that they should cooperate with the user tightly to figure out bottlenecks during the machine or line’s lifecycle and make constant improvement.

Bibliography

• [1] How to identify bottlenecks in a packaging production line, PackIot, December 20, 2019, https://packiot.com/bottlenecks-in-a-packaging-production-line/
• [2] Overcoming Manufacturing Bottlenecks: Tips for Improving the Flow of Production Patrick Lemay Dec 12, 2022, https://tulip.co/blog/overcoming-manufacturing-bottlenecks/
• [3] The impact of Industry 4.0 on bottleneck analysis in production and manufacturing: Current trends and future perspectives, Computers & Industrial Engineering, Volume 174, 2022, 108801, ISSN 0360-8352, https://doi.org/10.1016/j.cie.2022.108801
• [4] How to conduct a bottleneck analysis, Dave Westrom, MachineMetrics, Process Optimization, July 15, 2021, https://www.machinemetrics.com/blog/bottleneck-analysis
• [5] Constantin Hofmann, Tom Staehr, Samuel Cohen, Nicole Stricker, Benjamin Haefner, Gisela Lanza, Augmented Go & See: An approach for improved bottleneck identification in production lines, Procedia Manufacturing, Volume 31, 2019, Pages 148-154, ISSN 2351-9789