Big Data Engineering: Current Skills and Future PLM Considerations

Lionel Grealou Data, PLM, Talents Leave a Comment


Businesses demand more intelligent products and intelligent products demand data intelligence.

The typical challenges with data engineering (which obviously relate to the ‘science of data‘) include one or more of the following considerations:

  • Data flow: considering extract-transform-load (ETL), traffic patterns, reconciliation, integration, conversion, and downstream data consumption.
  • Data accuracy: removing data duplication and redundancy, optimizing master data management, through data cleansing, data aggregation, data diagnostic, dash-boarding and reporting.
  • System stability: taking into consideration data volume, variety, velocity, veracity, validity, etc.

Big data is data that typically does not fit into traditional relational databases, such as unstructured data or large analytics.

Big data involves taking multiple technologies and stitching them together to solve a business problem.

Big data engineering is a new field with a lot of new technologies and new positions which typically requires a subset of skills, such as data analysis (including statistical analysis, machine learning), data warehousing, data mining and collection, transformation, etc. – in order to spin big data into gold. 

If big data analytics could unlock new levels of profitability and efficiency, then surely data engineers held the keys. They combine both data engineering and data science. The reality is more about:

Data engineers help facilitate getting data from a variety of sources and in the right formats, ensuring that it adheres to quality standards, and that downstream users can access it quickly so that they can perform whatever required downstream tasks.

What is the relationship between Product Lifecycle Management (PLM) data and “big data”?

In the manufacturing space, the use of big data somehow lags behind other areas (…). Nevertheless, soon the benefits of big data could serve the entire manufacturing value chain, mainly in research and development, supply chain management, manufacturing, service, and other steps, helping reduce the development cycle, optimizing assembly process, increasing yields, and meeting customer needs. 

Big data has the potential to bring huge benefits to the manufacturing sector, such as described by Jingran et al. (2015):

  1. Enhancing product design quality and innovation.
  2. Improving production output quality and accuracy.
  3. Providing accurate, high-quality, personalized product service.
  4. Accelerating the integration of enterprise IT, manufacturing, and operation system to enter the age of intelligent manufacturing or Industry 4.0.
  5. Accurately predicting product demands.
  6. Accurately predicting supplier’s performance.
  7. Providing manufacturing equipment with intelligent sensing, management and maintenance (ref. Internet of Things).
  8. Supervising and controlling energy consumption and discharge (ref. circular economy).

There are still many challenges ahead in terms of big data maturity in PLM, complex data mining and profiling, as well as entry barrier due to the level of investment required to break through engineering and manufacturing data, process and technology complexity and integration. The big data skills are there, but the manufacturing industry lags behind others sectors in leveraging modern data models and technologies (…).

For decades, manufacturers have implemented IT systems to manage the product lifecycle including CAD, engineering, manufacturing, and product development management tools, digital manufacturing, etc. (ref. periodic table of PLM for full scope, and a brief history of PLM). However, the large data-sets generated by these systems have tended to remain trapped within their respective systems, with limited integration strategies – other than perhaps tactical ERP-PLM interfaces.

What are your thoughts?


Reference

  • Jingran L, Fei T, Ying C, Liangjin Z (2015) Big Data in product lifecycle management, The International Journal of Advanced Manufacturing Technology, 78(5-8).

This post was originally published on LinkedIn on 8 March 2016.