The analysis of data is one of the most important activities in modern asset management. It can make the difference between success and failure.
The Data Analytics in Maintenance course aims to create an appreciation of analytics in the context of a rapidly increasing data driven world. The South African mining industry was historically characterized conventional labour-intensive mines. Technology has been widely adopted and in 2021, most new mining projects are heavily mechanized and rely on various technologies, albeit with different levels of adoption.
This wave of mechanization has drastically increased complexity in terms of:
- supply chain systems: machines have thousands of parts and components
- training and skill development: the variety of assets are increasing fast
- cost: managing costs and operational efficiency has become more critical
- operational visibility: identifying operational bottlenecks in an environment with many complex coupled variables will need more methodical approach
These demands are not only applicable to maintenance aspects in mining, but also in any capital-intensive industries where equipment downtime and marginal changes in the operational expenditure can threaten the bottom line.
The course is a jargon-free practical guide to creating valuable and actionable insights to optimize the asset management effort. We have intentionally elected to create content that will accommodate learners with limited statistical training while limiting the primary tool of analysis to Microsoft Excel. The course will cover the whole analytics process using practical hands-on examples.
2. Course Duration
The material is taught in 5 days with a daily 8 hour per day schedule.
Learner should have a working knowledge of computers ( internet, Microsoft Excel and email).
4. Course Content
4.1 Module 1: A brave new world
Data analytics and concepts are introduced. The learner is exposed to how world-class maintenance professionals are using data strategically in business. The data science analytics process is explained and contrasted to how the course aims to address the various stages of the analytics process.
- The analytics process
- Business case studies
- The changing maintenance landscape
- Maintenance 4.0 technologies
- Organizational data analytics maturity
4.2 Module 2: Performance Measuring
This module discusses how and where an organization can begin its digitalization journey. The performance measurement is a natural starting process in starting up a concerted analytics process. The themes covered outline how an organization can effectively use performance measurements to identify risks, opportunities, foster innovation and drive continuous improvement.
- Maintenance and operational indicators
- How maintenance key performance indicators (KPIs) influence strategic outcomes
- How to make winning KPIs
- How to make winning KPIs
- How to dig deeper in the KPIs to understand what is driving them
- Understanding historical and predictive indicators
- Planning for effective performance measures
- Cascading strategic objectives through performance measures
4.3 Module 3: Data Sourcing
This module introduces the Extract, Transform and Load tool through Power Query. This tool is used to automate much of the data sourcing process and to a extend the data cleaning process. We use hands-on practical data to work on.
- Various types of structured data and how to digitalize them
- Data definitions through the knowledge pyramid
- Meta data management
- How to use Power Query to;
- Connect to various sources (online, folders, worksheets and database)
- How to transform data and automate the data preparation
- How to reduce manual operation when merging and appending data
- How to make regular reports to self-update and safe valuable time
4.4 Module 4: Cleaning Data
Most collected digital data that is not analysed is in the format that is not easy to manipulate. It could be some parts are missing, aggregated, the format is not usable to the analyst and no consistency in how the records are captured. This chapter gives the learner tools to make the data usable for the analytical process.
- Dealing with the top 5 data quality dimensions:
Practical examples on MS Excel and understanding functions to clean data
4.5 Module 5: Data Discovery and Analysis
Analysis is the heart of the analytics process. In this chapter we use methods to make the data to tell a compelling story. This module uses a practical examples where various data sources are integrated and prepared to solve real maintenance problems. Although many operations have some form of visualization reporting, the student is introduced to diagnostics and predictive analytics processes.
- Framing maintenance questions into analytical projects
- Creating focus to find maintenance insights
- Data aggregation techniques
- Creating visuals that tell a compelling story
- Going beyond descriptive analytics
- Diagnostic analytics and finding relationships within the data
- Prescriptive diagnostics: how to create data models
- Introduction to regression
- Time-series analysis
- Developing hypothesis
- Practical exercises to find meaning in otherwise latent data
4.6 Module 6: Trend Analysis Framework
This framework demands all the skills picked up from Module 3 to Module 5. It covers a novel method to track the consumption of 1000s of parts in a non-graphical method. This ingenious trending method can be used to answer questions such as: what is pushing the cost curve, which failure modes are on the rise and which parts must we reduce or increase in the warehouses
In an operation that uses 10,000+ individual coded items such as spare parts, raw material and fuel, an engineering manager is interested in understanding what is causing maintenance costs to rise in spite of many initiatives by the team. He ropes in an analyst to assist the reliability engineer to conduct a spend analytics project.
Tracking an individual item using time-series graphical plots may be adequate, however, when looking at 1000s of items, a graphical approach becomes rather tedious because;
- Creating 1000s of graphs is not practical to do or interpret
- Graphical information still needs further analysis to determine the trend
- Graphical information methods cannot be filtered or manipulated
This analysis framework opens up many possibilities in analysing trends in a powerful method.
4.7 Module 7: Formulating a Data Strategy
Data analytics is prevalent in some form in every enterprise. However, the maturity and benefits derived are not consistent. In the maintenance realm, there are countless amount of data generated within the CMMS system, the ERP and the production environs. This module shortly provides a brief overview on how an organisation as a whole or a maintenance department can take intentional steps to make the most of the data in the organization through crafting a Data Strategy. Aspects of the Data Strategy are discussed in a jargon-free manner. The discussions are aimed at maintenance professionals and not IT professionals.
- What is a data strategy
- Benefits of conceptualizing a data strategy
- Data strategy formulation steps
- Questions to create a data-driven operation vision
- Reviewing the current data utility
- A brief overview of what good data maturity looks like
- The four types of analytics
- Practical maintenance data considerations
- Data governance and how to leverage ICT models
- Creating a roadmap
- Typical Data Analytics jobs
- Typical Data Analytics software and tools
Who Should Attend
This course is intended for anyone who has to work with data and understand the fundamental meaning of data. These include Asset Managers, Maintenance Managers, Maintenance Engineers, Managers in control of Asset Management, Managers in control of Maintenance Stockkeeping and Purchasing, Maintenance Supervisors, Maintenance Planners, and many others. It can play an important role to prepare maintenance and operational personnel to understand the fundamental asset management/maintenance organisation's problems and enable them to manage and direct the organisation successfully.
Credits 12*, level 6**
* The course comprises 60 hours of study, of which 40 hours are in class, with a further 20 hours spent on completing a work related assignment..
**Higher Diploma level