Appunti completi di Industrial Management Toolbox

Toolbox

L. AS THINKING

The Most Underrated Skill in Management

• Ability to formulate a good problem statement
• Two ways our mind ends with problems
• Conscious process
• Focus effort demanding cautiousness
• Unconscious process
• Automatic pattern/habit based cautiousness

A Good Problem Statement

1. What is the problem important?
2. If we do not address? EAP, strategic now if it is not measurable
3. Narrow down to a specific miscalculation or a core issue (avoid big picture)
4. It references a situation, like agricultural cases (i.e., XYZ) and connect from equivalent to clear and specific goals
5. It contains a clear explanation of the gap between the current state and the goal
6. Key indicators are quantifiable: target, current state, gap
7. It is as neutral as possible concerning possible outcomes or solutions
8. It is sufficiently small in scope that you can tackle it quickly

Common Mistakes

1. Failing to articulate the problem statement
2. Ambitious plans and diagnoses of solutions
3. Lack of a clear gap people can't know
4. Problem is too big can’t treat high barriers

Structured Problem Solving

PDCA cycle

Using A3 form:

1. Formulate problem statement - Background section to explain why it is important and relevant with 5W1H
2. Document current design - Observe the work done (like you're authorized!)
3. Root causes - link observations to problem statement
• Hypothesis, 5S, Fishbone diagram
• Explain how the system generates the observed challenges (e.g., pit and routine accident)
4. Target design - update system to address the problem
• Specific, only related DO NOT JUMP TO THE END AND TEST POSSIBLE SOLUTION
• Improvement goals: necessity of the improvement (how much and how long?)
• Underlying problems: barriers or constraints that cannot be violated
5. Execution plan - plan for implementing proposed design with needed well-planned activities with an owner and a delivery date
• Track each activity after each task has been made
• Document what has been learned - results + opportunities

A3

A tool for supporting problem solving and process solving

A management approach to foster and develop a continuous improvement culture

The 7 elements:

1. Logical thinking process
2. Objectives
3. Results and process
4. Symptoms, distinction and prioritization
5. Alignment
6. Coherence within and across cases
7. Systemic visualization

General sections

1. Problem explanation
2. Problem breakdown - current situation
3. Target
4. Analysis of the root causes
5. Proposed countermeasures
6. Implemented countermeasures
7. Improvement results to pass
8. Standardize to share process

Background

• Fully clear your thinking about this problem
• 3 considerations required: implications for the company, how it’s linked to obj/KPI
• Understand who is the audience

Breakdown - current sitch

• Where do we stand?
• Go to the Genba (where the problem is)
• Collect data (compare with past/policy/depth/companies)
• Talk with people
• Check picture of the problem (flowchart)

Target

• Where we need to be?
• Clear gap
• Specific targets
• Nice to have - Not to have
• Specify time
• Quantitative terms

Root Cause Analysis

• Find the root cause(s)
• 5 Why(s)
• Sufficiency
• Process
• Time
• Mind on the achieved position
• Root cause (anomaly)

Countermeasures

• Based on the root cause(s) identified
• Keeping the mind process

Implementation

• Activities required
• Who is responsible
• Action on ground to be taken

Follow up actions

• What can I do next
• Which goal(s)_{importance of the actions}
• How to submit report needs?

Tips

• Visual part is important
• Fill one section at a time
• Remind who’s the audience
• Individuals in the breakdown of the problem
• Indicators in the target

CAUSE - EFFECT DIAGRAM:

Ideas (effects) to put in connection with all identified problems (causes), in order to identify correlation, relationships between causes and effects.

Keep the working team focusing on problem context, not only on life history of the problem, or on the point of view of each single team member.

Focus on causes and symptoms

• Define relationships among causes
• Define the level of detail
• Connect the causes which keep representing the knowledge that the team wants specific prob.

REGRESSION ANALYSIS (REGRESSION)

It identifies the relationship between two variables, allowing possibility to understand the effects caused by a controlled change of the corresponding value of the other.

• Dependent and many independent variables - relation?

- Single Regression:

1. Identification of the various method
2. Data collection
3. Effective representation of data -> scatterplot
4. Identify key variable and correlation factor (R)
5. Theory validation interpretation
6. Identification of the relation between variables

y = a + bx + e

PARETO ANALYSIS - PROPERTY INSET

• 80% of effect desc 20% of root causes
• Identify, apply to where just where I have less plt of the innovation
• Do not apply to much accelerated wheel not make the car accelerate

Classification -> Differentiation -> Allocation (of the action to four)

STEPS

1. Collect data -> eliminate outliers
2. Correlate which are the most import variables
3. Do a test for each value
4. Consistently use values in order to make premises
5. Actions for each class

Question to Ponder

• How much change in the crop is too small to detect?
• What is the assumption (cause of the crop)?
• Are there multiple outputs? Priorities?
• What is the objective: feeding, financial, ethical?
• Take output when estimating cost/acre

What Factors to Select?

Sources

• Literature
• Compromise/Effect/Known
• FGRA
• What does study suggest
• Avoiding knowledge
• Problem
• Embryonic knowledge
• Historic experience
• Rumsfeld's theory
• Opinion/Experience
• Observation
• Weather/Applied impact

Select based on if the factor is:

• Practical - Do you make sense to change their factor cause? Excessive cost/rant?
• Feasible - Is it physically possible to change the factor?
• Measurable - Can you measure (and report) the factor level settings?
• Low-hanging fruit
  • Quick experimental factor
  • Quick potential for significance on key measure
  • May go mad suit aimed to improve/lowering average

To deal with them

• Too expensive to control - We don't want to control/cannot be controlled
• A basic rule is to measure/control as much of it as you can and maximize what you can
• We cannot attempt to incorporate these factors into the design
• We can eliminate or move “to expect” an impact for these factors many fac
• We can attempt to hold those/each factor, constant during the cause of the experiment

What Level of the Factor to Test?

Common Numbers

• Levels set too close together - Here may be not enough variance in the response to adequate the experiment - high and low may get comforted and results not applicable
• Levels set too far apart

“Inter motives to control” - Assuming widest coir of significance access to combining conditions min-linear relationships might go undetected/creative problems

For a Screening Experiment

• Set “broad” excess - If we vary the factors to reasonable extremes, we will see their effect (if there is) - Risk: Many excessive variances, level pegging rather correlation

For a Characterization Experiment

• Precision/Accuracy is to learn from earlier experiments and adjust the factors accordingly
• Use the earlier known the methods, the spacing of the (eval) is generally reduced

For a Optimization Experiment

• Factors’ levels are set close together around invested level - Usually just influence mods