Accelerating Young Graduates’ Path to Computer Engineering

Author: Gawie Reyneke
Collection, Contextualization and Analysis Using Industry Standards
Introduction
Computer engineering is a critical field that involves the design, development, and optimization of computer hardware, networks, and software solutions. In any engineering discipline, prioritizing public safety, asset and equipment security, and environmental stewardship is essential. To help professionals achieve these objectives, a Code of Practice for Computer Engineering has been established. This code outlines specific computer engineering tasks, classifying them by complexity and sensitivity, and emphasizes key elements such as risk assessment, regulatory compliance, quality assurance, effective communication, professionalism, ethics, and lifelong learning.
The program equips candidates with both technical and soft skills, which are crucial for success in today’s modern workplace. It blends practical applications with theoretical knowledge to ensure comprehensive development. For this discussion, we will focus on one training component that is particularly important for a computer engineer working in mining and manufacturing environments: Manufacturing Execution Systems (MES).
MES systems are the backbone of modern manufacturing operations, ensuring that data from the shop floor flows seamlessly to business systems for real-time monitoring, control, and decision-making. For young graduates pursuing a career in MES systems engineering, mastering the key aspects of data collection, contextualization, and analysis is essential. This blog delves into these concepts, focusing on how data contextualization can be tied to industry standards like ISA-95 (the international standard for the integration of enterprise and control systems).
The four pillars of ISA-95 are critical to ensuring seamless integration and optimization:
- Production: The adoption of ANSI/ISA-95 regulations in this pillar facilitates better integration between business and control systems.
- Maintenance: Implementing ISA-95 guidelines in this pillar ensures efficient management of assets and resources.
- Quality: Applying ISA-95 to the quality pillar promotes uniformity in quality control processes.
- Inventory: ISA-95 guidelines improve inventory management efficiency in this pillar.
1. Data Collection: Laying the Foundation
Data collection is the first step in establishing a robust MES infrastructure. In an industrial context, equipment such as mills, motors, and gearboxes generate valuable operational data from sensors. Understanding how to effectively capture and manage this data is critical.
For example, in a mill system, which consists of a motor, gearbox, and cooling water circuit, the following parameters could be monitored:
- Vibration sensors: Measure vibrations in the motor and gearbox to detect imbalance, misalignment, or bearing wear.
- Temperature sensors: Monitor the temperature of the motor, gearbox, and bearings to ensure safe operating conditions.
- Oil pressure and flow sensors: Ensure the lubrication system is functioning correctly and providing sufficient oil to the gearbox.
- prevent overheating of the equipment.
- These sensors generate raw data, but without context, the data holds limited value. That is where ISA-95 plays a critical role by structuring the data in a standardized equipment hierarchy model.
To properly contextualize the data, the ISA-95 Equipment Hierarchy Model is used to standardize how manufacturing data is structured and organized. This model divides the equipment into different levels of abstraction, from the enterprise level down to the equipment and sensor level. Understanding this hierarchy allows engineers to systematically gather, store, and interpret data for operational insights.
In our mill system example, the data can be contextualized as follows:
- Enterprise Level (Level 4): Monitors overall production performance and connects to ERP systems for business management.
- Site Level (Level 3): Refers to the specific manufacturing site where the mill is located.
- Area Level (Level 2): Represents a specific area within the site, such as the milling section.
- Work Centre Level (Level 1): The actual mill system, including the motor, gearbox, cooling system, and lubrication system.
- Equipment Level (Level 0): Includes the physical components and sensors, such as the vibration sensor on the motor, the temperature sensor on the bearings, and the oil flow sensor in the lubrication system.
ISA-95 can be used to illustrate this hierarchy, showing how the mill system is decomposed into subsystems (motor, gearbox, bearings, lubrication, etc.) and sensors (vibration, temperature, pressure, flow).
By aligning the data from each sensor to this hierarchical model, engineers can contextualize the data to provide actionable insights. For instance:
- The vibration sensor on the gearbox indicates abnormal readings, suggesting potential wear on the gearbox bearings or inadequate lubrication.
- The temperature sensor on the motor shows rising trends, indicating cooling inefficiencies that need attention in the water or lubrication circuit.
Contextualization helps engineers correlate data with specific equipment and operational conditions, enabling them to make accurate and timely decisions.
3. Data Contextualization – Making Data Meaningful
Collecting data is only the first step; the true value of data lies in transforming it into useful information. Contextualization provides meaning to data by integrating it with metadata, such as time, equipment location, operational status, and equipment hierarchy from ISA-95.
Unified Data Management (UDM) can complement ISA-95 by ensuring that all collected data is managed centrally and consistently across different systems. While ISA-95 provides the framework for categorizing and contextualizing data, UDM ensures that this data is accessible, consistent, and unified across platforms, allowing for advanced analytics like anomaly detection and predictive maintenance.
In the mill system example:
- ISA-95 organizes and contextualizes sensor data, ensuring that each sensor reading is linked to the correct equipment and operational status.
- UDM manages the flow of this contextualized data across systems, making it available for predictive maintenance models.
For example:
- A vibration anomaly detected in the motor’s gearbox is contextualized by ISA-95, linking it to potential wear in the gearbox bearings.
- UDM then manages this data in real-time, allowing the MES system to trigger an alert and schedule preventive maintenance before failure occurs.
- Historical data managed by UDM can also be used to train machine learning models, improving the accuracy of predictive anomaly detection.
4. Next Steps: Building Predictive Anomaly Detection Models
Once data is collected, organized, and structured at the appropriate granularity, it lays the foundation for advanced techniques such as anomaly detection and process optimization.
In the mill system, for example, a predictive model can focus on anomaly detection for the motor, gearbox, and cooling system by using a combination of historical and real-time data. This model allows for the early detection of potential issues, such as misalignment or wear, enabling maintenance teams to address problems before they escalate into costly failures or downtime.
By applying these models, manufacturers can enhance operational efficiency, reduce unexpected downtime, and increase the longevity of their equipment. Predictive maintenance becomes a proactive strategy that continuously improves through machine learning and real-time data integration.
Conclusion
To accelerate the development of young graduates into proficient computer systems engineers, it is crucial to focus on training them in data collection, contextualization, and analysis. By utilizing industry standards like ISA-95, complemented by UDM, engineers can ensure that data is properly structured, accessible, and meaningful. This, in turn, allows for the development of predictive models that reduce downtime and enhance operational efficiency.
Incorporating real-world case studies, such as a mill system with multiple subsystems and sensors, provides a comprehensive understanding of how to apply these principles in practice. By mastering these techniques, graduates will be well-equipped to manage and optimize manufacturing systems, driving continuous improvement in the processing and manufacturing landscape.