My youngest son works at a liquid soap manufacturing facility in Cambridge, Ohio. Sunday evening I talked with him for our regular weekly call and asked about a project for which he is spending a lot of time. He mentioned that most soap factories mix all of the ingredients in large vats and then fill individual containers. He went on to say that his facility is more advanced in that it mixes the ingredients in the piping, does the appropriate cooking, mixes more ingredients, and eventually fills the individual containers. This is a much more streamlined approach where the only bottleneck is at the end.
I listened intently as I had no idea that making soap mirrored some of the data processes I work on at Sony. We have some data pipelines where all of the data is stored together and then cleaned in batches. Other pipelines clean the data as it comes into the system and quickly stores it in the tables that get used for analytics. The advantage of the second method is the data is always up-to-date. With batch processing you have to wait until the batch is processed before you can analyze the data. That may occur on a daily, weekly, or monthly frequency.
I explained the similarities to my son and he further elaborated that in the factory they have real-time monitoring of the systems. This includes flow rates and temperature values for different parts of the process. He asked if we have the same types of monitoring for the data and I responded that we do.
While my son is a Mechanical Engineer and I am an Electrical Engineer, the same ideas can be used in both disciplines with correspondingly similar pros and cons. This reminded me of a class I once took on "Thinking Outside the Box." The course mentioned that you may get ideas for solutions to problems you are working on by simply looking at other seemingly unrelated fields. I can't agree more.
No comments:
Post a Comment