These certainly have value since an end-to-end customer-focused business lean enterprise strategy requires careful evaluation from the perspectives of many disciplines. However sometimes when visiting a customer, supplier or potential acquisition there is only time for a brief gemba walk. In cases like these we must rely on what we can see for ourselves and summarize on a single A4 page in order to gauge the level of lean management. In order to do so without being misled by superficial signs of lean implementation or explanations of why things are they way they are, there are three things to check during a gemba walk: standardized work, kanban and hour by hour boards.
Being by checking standardized work. The reason is that poor or lacking standardized work may foil any attempt at a quick but meaningful assessment. When standardized work is poor the method may not be agreed, followed or repeatable, the inputs may not be correct, the sequence or combination of activities may vary, as will any the quality of one’s assessment.
Things to check:
- Is it properly documented, based on type of standardized work (1, 2 and 3)?
- Have these been revised within the past month?
- Is it being followed?
- Is there a way to immediately call attention (e.g. andon) when standardized work cannot be followed?
- Is the standard work in process inventory clearly and correctly identified?
- Does standardized work exist for the off-line, non-routine or occasional work?
A strong showing in the standardized work area is evidence of underlying stability, an understanding of takt – flow – line balancing, the limitation of motion and waiting wastes, the value of standardization as a prerequisite to a continuous improvement culture.
If standardized work appears to exist in a passable condition then next to check is kanban. We need to verify that the amount of work that is in the system is controlled, at an appropriate level and properly linked with information flow. If this is out of whack, being on-time 100% to a false demand based on overproduction will not mean much when it comes to the hour by hour performance.
Things to check:
- Is WIP limited rationally based on factors of demand, supply and variation?
- Is there a clear separation between regular and safety stock quantity?
- Is information visibly linked to the material (or in transactional terms does the “request” match the “work”)?
- Does every work piece or container have a kanban attached?
- Is there a regular and frequent collection and distribution cycle for kanban?
- Is the kanban quantity being reduced continuously?
A strong showing in the kanban category is evidence of the limitation of overproduction and inventory wastes, value stream thinking via linking of material an information flow, and the prerequisite leveling or reduction of lot sizes to the necessary level of flexibility to support a pull system.
Hour by hour boards
Once we are confident that the right work is being done in a standardized way we can check into how smoothly this system runs within a shift and within the hour. In addition, even a passing glance at a few hour by hour boards and their “Remarks” or “Reason” column will reveal the level of problem awareness and cadence of continuous improvement within on organization.
Things to check:
- Is there a plan for the day, shift, or hour?
- Is the lot size or batch size based on firm order quantity?
- Is the actual performance for each hour posted?
- Are reasons for deviation from the plan clearly specified each hour?
- Is there documented evidence of problem response within the hour?
- Is there an improvement target?
The vigorous use of hour by hour boards is evidence of strong shop floor management, early detection of problems, the control of quality and reduction of defects, short cycles of response to abnormalities and evidence teamwork across the organization.
The examples shown are from manufacturing and supply chain but the analogous tools should be visible in a lean service organization. All three of these lean systems are examples of visual management, setting standards and built-in action towards continuous improvement of safety, quality, on-time delivery and cost. When time allows, using open-ended questions (5W1H) rather than closed (yes / no) questions are preferable for deeper learning whenever the answer is “no”. Given an hour or two for a more thorough assessment, there are another 20-25 targeted questions to ask on each of these topics, with “Why not?” being one of the most important.