Siemens PLM: Cloud Services and Big Data

You can say that buzz around big data is annoying. At the same time, organization are struggling with a fundamental challenge – there are far more data than they can handle. Some interesting facts about data growth around us. Back in 2000, only 25% of all data stored in the world was digital. By 2007, 94% of all data was stored digitally. Some experts has estimated that 90% of all data in the world was produced for the last 2 years.

Manufacturing and engineering organization have to deal with a growing amount of data. Old fashion methods of handling data are not good anymore. You may want to look on some of my previous posts – Will PLM vendors will dig into big data? , Big data and importance of information lifecycle. Even more, the question of how to use data to improve product quality or design becomes important – PLM and big data driven product design. For many organizations data can become a very disruptive force.

Last week at PLM Connection 2015 conference in Dallas, I learned few interesting facts about how Siemens PLM is developing big data cloud solutions to handle large volumes of complex information for manufacturers. Steve Bashada’s presentation was about the work Siemens PLM did following the acquisition of Omneo, which was part of Siemens PLM acquisition of Camstar.

Getting back to Siemens PLM Omneo. The idea is to discover data patterns that can lead to optimal product performance. This is may sound too generic. However, if you translate it into more specific actions. Think about finding reasons why the last batch of hardware devices such as computer flash drive or wearable gadget was defective and track a supplier of faulty components. Inside Big Data whitepaper gives you an interesting perspective on Omneo solution. You can download whitepaper in exchange of your email address here. Here is the passage from the article I specially liked:

For a compelling example that illustrates how big data is affecting the manufacturing sector, we can consider Omneo, a provider of supply chain management software for manufacturing companies. The business need was to enable global manufacturers to efficiently manage product quality/performance and customer experience. Consequently, Omneo needed to collect, manage, search and analyze vast amounts of diverse data types, and it sought the right software and hardware infrastructure to support this effort.

• Enables global-brand owners to manage product performance and customer experience
• Delivers a 360-degree view of supply chain data
• Searches billions of data records in less than three seconds
• Scales to support 300 million records every month
• Allows customers to quickly search, analyze and mine all their data in a single place so that they can identify and resolve emerging supply chain issues

The following slide can give you generic yada-yada about the solution.

Siemens PLM is working on a solution with few selected customers. Dell is one of them. The following slides gives you an idea how a specific customer problem can be solved.

The solution uses “search based” user experience to search, filter and navigate between bits of data.

What is the technical foundation of the solution? Omneo is using some elements of existing big data stack you might be familiar with – HDFS, Hadoop, Cloudera combined with open source search technologies like SOLR. Omneo brings meta data and unified data model to handle product information and uses HBASE to manage information. The following slide can give you some more information about technical stack and how product is handling data.

What is my conclusion? Big data is a hard problem to solve. But it brings very interesting business cases. Siemens PLM Omneo is an example of specific data solution targeting big data problems in manufacturing organization. So far, the most specific example I was able to find reported by PLM vendors. My hunch, other PLM vendors might be looking on solutions, but haven’t seen specific publications about that. I think, big data can be applied in a very interesting ways to handle different product development, customer and manufacturing issues. We just not there yet. Manufacturing organizations and existing vendors are too slow to discover them. Just my thoughts…

Best, Oleg

picture credit Inside Big data article