Rules at the system's core
The person carrying out a fix cannot verify their own work. Domain rules are also reflected in database constraints.
03 / Web · Manufacturing
Gemba grew out of my work with mLean in manufacturing. In it, I connect findings from workplace walks with corrective actions and checks that the problem has been resolved.
EXPERIENCE IN LEAN MANUFACTURING
Implementing and using mLean helped me understand what I needed from a digital Lean and kaizen tool. Some features were missing or did not suit the way I worked. That led me to create Gemba.
Someone records a problem and adds a photo or a note. An assignee takes ownership and provides evidence of corrective action; another person verifies the result. Walk schedules, work standards and operation measurements connect individual findings with everyday manufacturing work.
A nonconformity becomes a finding. A photograph or note adds detail about what was observed on site.
An assignee and a due date define the next step. The action needs supporting evidence.
Another person verifies the fix. Recurring and safety-related issues also need an effectiveness check.
BROWSER SIMULATION
Try selected workflows directly in your browser.
This is a separate simulation using fictional data.
Loading the demo… A text overview is available below.
TECHNICAL DECISIONS
The person carrying out a fix cannot verify their own work. Domain rules are also reflected in database constraints.
An audit uses a frozen version of its template. Later changes must not overwrite that basis.
Go, SQLite and HTML in one system. Core features and source retrieval do not depend on a language model.
A rejected transition must neither change the state nor create a false audit event.
Template versions, scoring schemes, recurring schedules and roles preserve the meaning of records over time.
Documentation supports full-text search with citations even without a model. People record decisions and audit answers.
Design and development of my own tool, drawing on my work with Lean methods and mLean. The focus is on connecting findings, responsibility, evidence and verification of corrective action.
The project has a demonstration environment with five fictional plants. This website presents selected workflows from it. Finding counts and photos are demo content, not results from a customer deployment.
The sources are the repository and a signed-in review of five demo plants on 18 October 2026. This web simulation uses selected text and generated photographs from fictional demo data. The displayed counts are not results from a customer deployment.
CONTACT
If a project caught your eye or you are working on something similar, get in touch. I would be happy to exchange ideas and experience, or discuss working together.