AI and Machine learning buzz is going through the roof. The number of new companies in the this categoryhas grown exponentially over the past few years. Few weeks ago, I shared my thoughts about AI-zation of CAD and PLM . Check out here. I found the topic confusing with too much buzzwords and marketing messages.
One of my conclusions in this article was related to the difficulties to find data hidden behind firewalls, data formats and legacy data management systems. Without such data future of machine learning in CAD and engineering system is questionable.
While AI is very much overarching term, one which in my view requires deeper understanding is machine learning.
Daniel Tunkelang, which I know back to his Endeca days, published a great article giving you simple and easy understanding of what machine learning is about. Navigate here to read – 10 things everyone should know about machine learning.
The article is short and sweet. It is must read. My favorite passage is the following one.
Machine learning is about data and algorithms, but mostly data. There’s a lot of excitement about advances in machine learning algorithms, and particularly about deep learning. But data is the key ingredient that makes machine learning possible. You can have machine learning without sophisticated algorithms, but not without good data.
If you plan to get involved into machine learning activity, this is a great reminded about what is the hardest task.
Most of the hard work for machine learning is data transformation. From reading the hype about new machine learning techniques, you might think that machine learning is mostly about selecting and tuning algorithms. The reality is more prosaic: most of your time and effort goes into data cleansing and feature engineering — that is, transforming raw features into features that better represent the signal in your data.
It made me think about PLM changing paradigm and data management. To make a difference new PLM platforms have to think about expanding their data collection processes to be able to provide a representative data sets to train future decision support systems. Web scale and machine learning will be an activity to support future leapfrog in product development and decision support. Without that, CAD and PLM systems role will be limited to controlling file vaults and release processes inside of organization. It is far from what manufacturing companies are demanding these days.
What is my conclusion? Rethinking data management paradigms from local to web scale. Future success of machine learning in product development, manufacturing and supply chain will come from web scale data collection. Global data management combined with multi-tenant engineering and production systems will become a skeleton of future business models in manufacturing software. Just my thoughts…
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.