PLM Big Data Heaven: Expectations and Reality

Manufacturing companies are learning about the value of data assets. For the last few years, we have seen a tremendous interest in data capture projects with the goal to make it valuable for manufacturing companies, consumers and engineering software vendors. It sounds like manufacturers are literally sitting on big data dynamite of potential revenues and opportunities driven by data initiatives.

There are several reasons that manufacturers can become a primary beneficiary of big data boom. Manufacturers are uniquely positioned to benefit from big data. Every industry and individual is touched by manufacturing. Manufacturers were among first industries to make wide data collection in a standard practice. Many examples from automotive, aerospace and other industries. Manufacturing companies typically don’t face the data collection barriers. Whether they know it or not, many consumers readily provide valuable data to manufacturers on a daily basis. The opportunity driven by big data can include improving product quality, help to discover new design for existing products and find new product opportunities.

It might sound like manufacturing is Big Data Heaven. The past 4-5 years has been like a dream for companies working on analytics and big data. Nevertheless, when I had a chance to speak to companies about Big Data projects in manufacturing, I’ve heard that reality is a bit different from expectations. Big data projects are overspending in data extraction and many other tasks needed to prepare data for analysts. Real big data projects are still very rare and requires massive software skills in both data processing as well as DevOps. Many projects are not going much future than setting up Hadoop cluster with some half-automated scripts.

In my blog last year, I shared thoughts why big data is a big problem for PLM vendors. Old data management technologies preventing data extractions, limited scale and openness are main factors preventing PLM vendors to into big data domain. Big data projects are using separate infrastructure and tools – unfortunately creating another data silo and disconnected projects.

It made me think about how to improve the outcome of big data projects in manufacturing.

1. Evolution of existing PLM tools into platforms enabling big data solutions. Existing big data tools such as Spark and Hadoop can provide good infrastructure for big data projects, but these tools are not user friendly and requires lot of customization to run in end user environments.

2. Development of practical manufacturing user cases and best practices for big data projects. These user cases can map manufacturing data into typical set of queries and data collections that can be reused by several manufacturing companies of the same industry.

3. Improvements of automation tools to optimize data hand-off and processing between system, databases, environments and other tools. New automation tools can provide specific self-services for manufacturing big data projects to improve productivity and deployment alongside with existing PLM and ERP infrastructure.

What is my conclusion? For most of manufacturing companies, big data is a big dream and research project. In order to turn big data project into reality, companies should develop environments to support continuous analytics and not one time huge effort to collect data and make it usable. These environments need to be connected to existing data management and data processing infrastructure. It will create an automated pipeline between data scientists, to big data engineers and finally product engineers working on specific big data solutions for manufacturing industry. Just my thoughts…

Best, Oleg

Want to learn more about PLM? Check out my new PLM Book website.

Disclaimer: I’m co-founder and CEO of openBoM developing cloud based bill of materials and inventory management tool for manufacturing companies, hardware startups and supply chain. My opinion can be unintentionally biased.