PDCA - Learning Along the Way
Upcoming SlideShare
Loading in...5
×
 

PDCA - Learning Along the Way

on

  • 13,148 views

We often miss the important role of prediction error.

We often miss the important role of prediction error.

Statistics

Views

Total Views
13,148
Views on SlideShare
6,960
Embed Views
6,188

Actions

Likes
7
Downloads
533
Comments
1

12 Embeds 6,188

http://www.lean.org 5831
http://leanfreaks.wordpress.com 275
http://test.lean.org 35
http://www.sehen-lernen.com 10
http://www.linkedin.com 9
http://10.1.1.8 9
http://sehen-lernen.com 6
http://10.24.41.136 5
http://www.encob.net 4
https://si0.twimg.com 2
http://10.200.161.8 1
http://translate.googleusercontent.com 1
More...

Accessibility

Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
  • One thing missing is when humans are in the uncertainty
    arena, they tend to operate from the fight/flight/freeze part of the
    brain. That's why we get resistance to any change effort. And, that's why this stuff can work, because the antidote to that is to engage the people affected by the change to help design the change!
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment

PDCA - Learning Along the Way PDCA - Learning Along the Way Presentation Transcript

  • PDCA LEARNING ALONG THE WAYMike Rother &Bill CostantinoAugust 2012 Illustration from: The Adventures of Ned the Neuron www.kizoomlabs.com Copyright © 2013 Mike Rother, all rights reserved 1217 Baldwin Avenue / Ann Arbor, MI 48104 USA / tel: (734) 665-5411 / mrother@umich.edu © Mike Rother TOYOTA KATA 1
  • WHY THIS SLIDESHARE? There are many challenges humans face, but we can handle them... if we manage ourselves a little differently Why should we manage ourselves differently? Because a scientific, iterative way of thinking & acting is not the typical way adults nor our business organizations think & act. We naturally prefer to define the steps weʼll take, determine whoʼs responsible for each step, assign Bill timing and execute the plan. Mike We do need to make plans, but the approach we may naturally prefer is often not effective for meeting goals in complex situations. It takes a different approach and a little practice to mobilize our astonishing collective capability for meeting challenging goals© Mike Rother TOYOTA KATA 2
  • TAKEAWAYS One of the ways we learn is through the steps of Plan --> Act --> Evaluate. This is a natural cycle. These steps are different from the mechanistic model that business is currently comfortable with. Plan --> Act --> Evaluate is mirrored by the PDCA cycle. When a step along the way goes differently than predicted or planned you often learn something new that helps you reach the goal. Although this is different from what business is currently comfortable with, we can practice a way to experiment in the “PDCA Zone.” Practicing the Improvement Kata allows us to actualize the PDCA approach in business and everyday life.© Mike Rother TOYOTA KATA 3
  • SO LETʼS TAKE A LOOK© Mike Rother TOYOTA KATA 4
  • HOW DO YOU WORK TOWARD A GOAL? Viewed from a distance, any human endeavor seems to involve three steps: Plan Action Evaluate© Mike Rother TOYOTA KATA 5
  • BUT OF COURSE IN REALITY YOUʼRE CONSTANTLY CYCLING THROUGH THOSE THREE STEPS When you try to reach any objective you repeat the steps many times P P P P P P P P P A A A A A A A A A E E E E E E E E E Why? Because we canʼt predict the future. No plan we make is 100% correct and nothing goes 100% according to plan. Reaching an objective involves learning and making adjustments along the way.© Mike Rother TOYOTA KATA 6
  • A LEARNING PROCESS IS GOING ON AS YOU TRY TO MOVE TOWARD A GOAL Take the example of walking toward something. Constant sensory feedback and subconscious muscular adjustments are happening in order to generate effective locomotion. While you walk, information from various sensors is used to adapt your posture and walking pattern to the dynamic characteristics of the task, the environment and your body.© Mike Rother TOYOTA KATA 7
  • HOW DO WE LEARN? Believe it or not... An essential mechanism of learning is Prediction Error What do you (Plan) expect to happen? Prediction Evaluate Action What actually happened? What did you learn? A main way we learn new things is when an actual outcome differs from the predicted outcome© Mike Rother TOYOTA KATA 8
  • HOW PREDICTION ERROR DRIVES LEARNING Itʼs the scientific approach: When a result is as-predicted it confirms something you already thought. When a result isdifferent than predicted you are about to learn something new. Strengthens Confirmed current thinking. Like re-walking a Prediction path in the snow. Surprise. Potential for new knowledge, Error learning & discovery. Prediction confirmation keeps you in place. Prediction error leads you out of your assumptions and forces exploration. This is because prediction error reveals a knowledge threshold.© Mike Rother TOYOTA KATA 9
  • THIS IS EVEN HOW SOME OF YOUR BRAINʼS NEURONS WORK Learning goes on all the time in your brain. Every time you do something the synaptic connections between the involved dopamine neurons may be strengthened (+) or weakened (-) based on Reward Prediction accuracy (+) or inaccuracy (-) Reward Prediction Strengthen Actual or weaken result Rapid adaptation to reality this one synapse Synapse (learning) via a cycle of prediction, met or not, evaluated by dopamine. Neuron Neuron© Mike Rother TOYOTA KATA 10
  • PREDICTION --> ACTION --> EVALUATE This cycle is a natural building block of anything that humans achieve as they operate in systems P P P P P P P P Big A A A A A A A A E E E E E E E E Project P P P P P Landing an A A A A A E E E E E Airplane P P P Driving A A A Plan E E E to Work P A Cooking Evaluate E an Egg Action Synapse© Mike Rother TOYOTA KATA 11
  • PLAN DO CHECK ACT PDCA (or PDSA) is a 4-step learning cycle that echoes our neural learning process Wow... a version of PDCA may be baked into our neurons that are involved in learning new behavior!PDCA drawingby Jurgen Appelo © Mike Rother TOYOTA KATA 12
  • HOW PDCA WORKS The “C” of PDCA The “P” of is a reflection... PDCA is an expectation or What are we a prediction... learning from this? ...a hypothesis What do we need to adjust? Illustration from The Team Handbook, page 3-33 Unexpected results redirect your thinking, forcing new interpretations and steps. They put you at the learning edge. When you reflect and attempt to understand why your prediction was inaccurate you discover new insights and build new knowledge.© Mike Rother TOYOTA KATA 13
  • PDCA EXPERIMENTS ARE DONE AT KNOWLEDGE THRESHOLDS Thereʼs always a knowledge threshold... look for it The knowledge threshold is the point at which you have no facts & data and start guessing. This is where you should do the next PDCA experiment; learning where the facts run out. Predictable Zone Uncertainty / Learning Zone tacl e s ? Ob s ? Next Target r Condition nclea ryThe plan is Current ? U ito Terr We wantmade here Knowledge to get Threshold here© Mike Rother TOYOTA KATA 14
  • FOR EXAMPLE We know how a pull system works, but we donʼt know in advance everything it will take to make your pull system work. This means you canʼt implement a pull system. Youʼll Target have to experiment your way forward and iteratively Condition learn how to make your pull system operate as desired.Predictable Zone Uncertainty / Learning Zone Pull system n g rimenti between Expe processes A & B working as designed,The plan is by (date) Currentmade here Knowledge Threshold Spot the knowledge threshold and conduct your next PDCA experiment here as quickly & cheaply as possible!© Mike Rother TOYOTA KATA 15
  • OUR NEURONS MAY UNDERSTAND THE ROLE OF “PLANNING” BETTER THAN WE DO Many of us think finding the best path to a goal involves developing the right plan and then executing it. Turns out, thatʼs incorrect when youʼre operating in an interconnected system. The way it works is that you make the best possible plan, and then you adjust along the way based on what you are learning along the way. If you go to business school and learn a planning process, thatʼs only half of the matter. You should also learn a good iteration process. Imagine what would happen if the neurons in your brain only planned and executed and didnʼt constantly adjust based on micro results. You probably wouldnʼt be alive reading this today! A plan is only a prediction of how things will go. A plan is a hypothesis.© Mike Rother TOYOTA KATA 16
  • WHAT ARE WE TEACHING? Too much certainty Human capability, passion-driven endeavor and iterative / adaptive thinking are alive and well on the planet. But itʼs often overshadowed by an unscientific kind of thinking that has become habitual in many large business, financial, political and academic organizations. Complex Interconnected System AMechanical System H BA B C D G C F D In business schools we teach how to manage E businesses as if they were mechanical systems, not the complex, interconnected systems they actually are.  In particular we teach using accounting and financial control methods to make predictions and strive for financial results. Unfortunately, we have a tendency (certainty bias) to put too much faith in our predictions, and financial results are only an abstraction of reality.© Mike Rother TOYOTA KATA 17
  • WHAT IS SCIENTIFIC THINKING? What we • Quantification and precision may think • Objective and certain scientific is • Reveals what is there Example: We have made the right plan • Involves uncertainty, What ambiguity & incompleteness scientific • Never free from error really is • A process of discovery, via systematic trial and error Example: Our plan is a hypothesis© Mike Rother TOYOTA KATA 18
  • WHAT SHOULD WE DO NEXT?© Mike Rother TOYOTA KATA 19
  • PRACTICE A DIFFERENT WAY! We can practice routines (kata) to develop new habits that take us beyond mechanistic thinking, to iterative thinking thatʼs more in line with how interconnected systems work Illustration by Meryl Runion Rose© Mike Rother TOYOTA KATA 20
  • ONE STEP TO PRACTICING ITERATION Whenever youʼre working toward a goal ask yourselves the Five Coaching Kata Questions daily PRESCRIPTION 5 Basic Coaching Kata Questions 1. What are we trying to achieve? 2. Where are we now? 3. Whatʼs currently in our way? 4. Whatʼs our next step (the next experiment) & what do we expect? 5. When can we see what weʼve learned from taking that step?© Mike Rother TOYOTA KATA 21
  • ANOTHER STEP TO PRACTICING ITERATION Conduct experiments inside the PDCA Zone Businesspersons are often understandably averse to experimenting because it feels uncontrolled; like you may never reach the goal. By using the Improvement Kata you can diminish this fear. Your Target Condition has a hard achieve-by date and is measureable. There are budget constraints and quality parameters. Itʼs within these defined limits that you design and conduct rapid successive experiments to reach your target condition.© Mike Rother TOYOTA KATA 22
  • FOR GOOD ORGANIZATIONAL HEALTH... Take a daily dose of Improvement Kata practice Bill Mike Practicing the Improvement Kata is perhaps the best way weve found so far for actualizing PDCA in an organization. ~ John Shook, Chairman and CEO, Lean Enterprise Institute© Mike Rother TOYOTA KATA 23
  • Remember... any plan is only a hypothesis, so be prepared to learn on the journey!© Mike Rother TOYOTA KATA 24
  • I once believed that PDCA thinking is a natural phenomenon, like the way the brainʼs dopamine neurons behave. But I also see all the evidence in front of me which points out that PDCA thinking is not as natural, automatic and widespread as I wish it were. Just because our brainʼs dopamine neurons utilize prediction error to learn doesnt mean that we as humans have learned to utilize prediction error. It takes practice.© Mike Rother TOYOTA KATA 25
  • WEʼD LIKE TO THANK...  Pat Boutier  Emiel Van Est  Dennis Gawlik  H. Thomas Johnson  Jeff Liker  Bernd Mittelhuber  R.R. ... for their thoughtful and insightful input to this SlideShare© Mike Rother TOYOTA KATA 26