Good information is key to make great business decisions.
But what is information exactly ?
Everyone knows information can come from data. While data is raw, believe it or not, information is informative. And the definition of informative is subject to the person consuming this information.
A technician debugging a faulty unit or machine will find the raw, text, testbench logs informative. He will search and comb through the text lines based on his experience to find an error or cue about why the process is misbehaving or why the unit is supposedly faulty.
So clearly raw testbench logs can be considered informative to a technician.
But, as is, these logs have very limited usage outside debugging. Would a production manager use testbench logs in his day-to-day decision-making ? Probably not.
However, what if the same logs were somehow aggregated to convey higher-level information, such as the rate of anomalies, the yields of certain sub steps or the most common failures ? Then these metrics may become informative to the production manager and even the director of operations !
Providing insights such as:
This month, the first-pass yield of calibration was up to 98.1% from 97.4% last month
This batch of production has 87% more failures of type 'XYZ'
All from the same, raw, testbench logs.
We haven't changed the available data. We transformed this data into new, useful information. The tools and processes that take you from data to decisions are usually referred to as a data analytics stack.
If good information is key to great decisions, then a well-built analytics stack is key to good information.
And to have good information, we strongly believe at TackV that your analytics stack needs to be automated, centralized, documented and tested. Data analytics automation.
That’s what we do. We build analytics stacks specifically for the challenges of manufacturing data analysis.
Knowledge on how to build analytics stacks for the web and marketing industries is plentiful, but not so much for manufacturing. We wish to provide companies the required resources so that they can build their own while avoiding common pitfalls.
Therefore, we will be sharing the tools and the process we use to build analytics stacks throughout a series of blog posts within the next couple of months. Stay tuned !