Collecting factory equipment signals and building the store for them

CoreData EngineeringData Platform EngineeringData Catalog & Metadata ManagementBI & Data VisualizationEmbedded & IoT

Organisation: YAMATO KOGYO Co., Ltd.
Period: May 2022 – Dec 2024
How it started: Assigned
My part: Design through build and operation, alone
For whom: The manufacturing division (steelmaking)
What changed: Equipment signals vanished where they were produced → Stored in a time-series database and analysable against individual products
Scale: 22 tables and over 450 fields in the integrated database, and over 1,000 time-series channels (as of May 2023; by the time I left, several thousand channels and a few hundred tables)

What I did

Designed and built the database the operational analysis needed, and wrote the programs feeding it from the business and plant systems.
Ran the time-series collection platform for plant equipment signals, including managing what is collected, watching the collection rate, and detecting bad or missing data.
Built a data catalog web app for listing and searching the collected items.

Semi-automated the configuration of the collection gateway, with scripts and a procedure, so the set of signals could be changed freely.
Designed the link between per-process records and the time series: deriving a time range from the cast length and the cutting record of the continuous caster, so the signals can be traced per product.
Built the signal list and the catalog web app in React and TypeScript.

Extraction that goes inside the implementation, visualisation that serves the business goal.
I work alongside the business from the definition of the metric itself, and stay with it through to the insight it was asked for. From IoT systems to web and mobile, I have broad experience of building software, so I extract data with a deep understanding of the implementation behind it. Visualisation and analysis are the same — I put generative AI to work and deliver quickly, in the form the person will actually use. Underneath that, I can shape the data, the context around it and the ETL that feeds it for the age of generative AI, so that people and AI alike can work with what is there.

I can design the whole system, not one layer of it
From hardware up to the cloud, I can draw the blueprint. I have built IoT hardware, web applications, mobile apps, and a game service platform running in the cloud, so I can combine pieces I have not used before and still see how the whole thing fits together. Holding that picture is also what lets me hand implementation to generative AI and still carry a whole system on my own.
I read the code, then deliver the data
Table definitions rarely tell you what a column means to the business, so I read the specification and follow the code until I know. Years of building software myself means I can open a running system, watch how it behaves, and produce data — or answer a question about how that data comes to exist — on that basis. It changes how I use generative AI for the same work: because I know where to look, I can direct it precisely, get an answer quickly, and judge whether the answer holds.