Manufacturing Process Analysis: Starting Points for Fast
Improvements
Submitted to:
Mam Adeela Pervaiz
Submitted by:
Group 4
Muhammad Usama Umar (09)
Murad bin Khalid (32)
Ibrar Hussain (43)
Shabahat Hussain (45)
Dilawar Shah (49)
Semester:
8th Regular (2020-2024)
Department:
Malik Firoz Khan Noon Business School
University of Sargodha What Is Manufacturing Process Analysis?
A manufacturing process's performance
is analyzed using manufacturing process analysis, or MPA. This approach involves examining procedures to make sure they satisfy the organization's requirements and are in line with its goals. These requirements can be broadly categorized into several key areas:
1. Quality Standards: Ensuring the products manufactured meet specific
quality benchmarks, which could include durability, reliability, functionality, and appearance.
2. Production Efficiency: This involves maximizing productivity through
optimizing the use of resources (such as materials, equipment, and labor) and minimizing waste. Efficiency requirements might include increasing throughput, reducing cycle time, and improving workflow.
3. Cost Effectiveness: Keeping production costs within budget without
compromising quality. This could involve strategies for cost reduction in materials, labor, energy, and overheads.
4. Compliance with Regulations: Adhering to all relevant legal and regulatory
requirements, including environmental regulations, health and safety standards, and labor laws. This often includes compliance with industry- specific guidelines and practices.
5. Environmental Impact: Implementing processes that reduce environmental
degradation, such as reducing emissions, minimizing waste production, and using sustainable materials. MPA can cover any stage of the product life cycle, including:
In the product design stage, MPA focuses on ensuring that
products are designed not only for functionality and aesthetic appeal but also for manufacturability and ease of assembly.
Assembly:
During the assembly phase, MPA looks at the actual construction of the product
Quality Control:
Quality control is critical in ensuring the product meets or exceeds
customer and regulatory standards.
Scheduling:
Effective scheduling is crucial to ensure that production processes flow
smoothly and efficiently
Maintenance:
Maintenance of equipment and facilities is essential to prevent
unexpected downtime and maintain consistent production quality To enhance smart manufacturing procedures, machine learning (ML) and artificial intelligence (AI) technologies are frequently combined with MPA.
The majority of the industrial industry is moving toward smart manufacturing.
Recent research indicates that until at least 2028, the machine learning industry is anticipated to develop at a rate faster than the 38.6% compound annual growth rate (CAGR).
When AI and ML are used with manufacturing process analysis, more insight may be obtained with less work. Faster advancements and more flexible corporate processes are made possible by effective big data analytics. Any process's efficiency may be rapidly evaluated, bottlenecks can be found, and solutions can be found using machine learning tools.
Machine learning systems have the ability to create and evaluate models to enhance several aspects of business, such as customer happiness, sales, and product quality, in addition to production capacity
How to Develop an MPA Strategy?
A plan is needed to use AI and ML
technologies to analyze industrial operations. Manufacturing process analysis typically involves the following steps in its workflow:
• Value identification • Data acquisition • Data processing • Data analysis • Machine learning • Model deployment
1. Value Identification
This initial phase involves pinpointing specific goals and areas where AI and ML can add value to the manufacturing process. It’s about understanding what problems need solving and what improvements are sought. This could involve:
• Increasing efficiency: Identifying bottlenecks in production lines or areas
with high downtime.
• Enhancing quality: Detecting defects earlier in the production process or
reducing scrap rates.
• Reducing costs: Optimizing resource usage, including materials and energy,
or minimizing labor costs through automation.
• Predictive maintenance: Identifying equipment likely to fail and scheduling
maintenance to prevent downtime.
2. Data Acquisition
Once the targets have been identified, the next step is to gather data relevant to those goals. This involves setting up sensors and data collection systems across various parts of the manufacturing process. The data could include:
• Operational data: Machine speeds, feed rates, and settings.
• Environmental data: Temperature, humidity, and other environmental
conditions affecting production. • Resource data: Consumption data for materials, energy, etc. Data acquisition should ensure completeness, accuracy, and timeliness to facilitate effective analysis.
3. Data Processing
Data collected from industrial operations often comes in raw forms that are not immediately suitable for analysis. Processing is necessary to clean, normalize, and structure this data. Steps include:
• Cleaning: Removing outliers, correcting errors, and handling missing values.
• Normalizing: Scaling data to a uniform scale to prevent biases based on
measurement scales.
• Feature engineering: Creating new data points (features) from raw data that better represent the underlying processes for predictive models.
4. Data Analysis
This stage involves exploring the processed data to uncover patterns and insights. Techniques used might include:
• Statistical analysis: Descriptive statistics to get a basic understanding of
data distributions and relationships.
• Exploratory data analysis (EDA): Using visualizations like histograms,
scatter plots, and box plots to find patterns, trends, and anomalies in the data.
5. Machine Learning
Based on insights and patterns discovered during data analysis, machine learning models are developed to predict outcomes or classify data. This step includes: • Choosing a model: Based on the problem type (regression, classification, clustering).
• Training the model: Using historical data to train the model to recognize patterns and make predictions.
• Validation: Testing the model on a separate dataset to ensure it performs
well on unseen data.
6. Model Deployment
The final step involves implementing the machine learning model into the production environment so that it can start providing real-time predictions and insights. Deployment includes:
• Integration: Ensuring the model interfaces smoothly with existing
production systems and data flows.
• Monitoring: Continuously checking the model’s performance to catch any
degradation over time.
• Updates: Refining and retraining the model with new data and feedback to maintain its accuracy and relevance.
Companies have to evaluate a process's worth and the possible benefits of any changes first. A manufacturer could wish to lower the quantity of waste produced by a certain process, for instance. It is important to monitor waste reduction, cost savings, and time savings using well-defined measures. MPA requires collecting data from each process element. This might include labor, cost, and total production statistics, as well as data from sensors attached to industrial machinery.
After then, the data is processed. To make sure the data collection is free of duplicates and missing information, it must be cleaned. Several analytical and machine learning techniques may be applied to produce insights once the data has been collected and cleaned.
Using MPA to Improve Product Quality
Enhancing product quality is only one of
the numerous benefits of incorporating MPA into decision-making. Finding problems that might affect a product's quality is made easier with a thorough examination of the production process.
By identifying the underlying cause of a quality problem, the analysis's knowledge
enables engineers and analysts to put a solution in place. With the correct tools, manufacturers can solve quality concerns more quickly.
Increasing Overall Equipment Effectiveness
The effectiveness of the equipment
being utilized has a big impact on both process cost and efficiency. MPA contributes to a higher return on investment (ROI) and improved production quality by raising overall equipment effectiveness (OEE). Machines with higher efficiency require less maintenance and have less downtime. They also have a lower probability of producing flaws. These advantages enable businesses to extract greater value from their production processes.
Challenges of Implementing Manufacturing Process Analysis
Implementing an MPA strategy includes
three key challenges:
• A high volume of data
• Wide variety of data • Access to real-time data
Access to a large amount of data may be necessary for manufacturing process
analysis. Data may be gathered by manufacturers from sensors, suppliers, distributors, and other outside sources.
Additionally, a wide range of sources contribute to data collecting. The information
might be kept in many physical places, cloud networks, and databases.
The capacity to handle data rapidly is further limited by the amount of diverse data that must be gathered and analyzed. With conventional data gathering techniques, manufacturers are unable to acquire real-time analytics. The secret to obtaining insightful and useful information is to use a specialized solution for data collection, processing, and analysis. For an MPA strategy to be successful, selecting the appropriate partner is also crucial. In order to acquire data, analyze it, and create better procedures, collaborate with a seasoned data science firm to put the systems in place.
