SlideShare a Scribd company logo
USING SIMULATIONS AND AI TO OPTIMIZE SUPPLY CHAINS
Illustrated using the Beer Distribution Game
Berlin, 28.5.2020
Transforming Enterprises
TOPICS TODAY
Introduc on to the Beer Distribu on
Game, a supply chain simula on.
Use computa onal modelling to
understand supply chain dynamics and
nd better playing strategies.
Train autonomous agents to play the
game using a machine learning /
reinforcement learning approach
DOWNLOAD THE CODE AND COMPUTATIONAL NOTEBOOKS
Download all
the code and
in-depth
computational
notebooks from
.GitHub
All resources are available via www.transforming-enterprises.com
THE BEER DISTRIBUTION GAME
An Introduction
The Beer Game was developed in the 1960s at MIT to illustrate how difficult it is to manage dynamic
systems – in this case a supply chain that delivers beer from a brewery to the end consumer
The game became well known in the 1990s after Peter Senge's description of it in his world-wide
best selling book The Fifth Discipline
The game is great because even though it is about a very simple system and despite very simple
rules, the resulting behaviour is quite complex.
–
–
–
A BRIEF WALKTHROUGH OF THE BEER GAME
Try it yourself at beergame.transentis.com
The game is usually played with four players, but the single player version is also fun.
But let's look at the rules and some
poten al pi alls rst!
THE SITUATION
You are part of a supply chain that delivers beer from a brewery to the end consumer
try to meet the demand of your respective customer at all times, while keeping inventory low.
–
–
THE RULES
The rules of the game are simple – in every
round you perform the following steps:
1. Check deliveries. Check how many units
of beer are being delivery to you from
your supplier in the supply chain.
2. Check orders. Check how many units of
beer your client in the supply chain has
ordered.
3. Deliver beer. Deliver as much beer as
you can to satisfy demand (the game
does this for you).
4. Make an order decision. Decide how
many units of beer you need from your
supplier to keep your inventory stocked
up.
1
2
3
4
SOME PITFALLS TO BE AWARE OF
Backorder and inventory costs
Delays
HOW PERFORMANCE IS APPRAISED
Individual Supply Chain Cost.
Your accumulated cost should remain below $8.300.
Overall Supply Chain Costs.
Total supply chain cost should remain below
$29.300.
LET'S GIVE IT A GO
https://beergame.transentis.com
ORDER BEHAVIOUR IN A TYPICAL GAME
What the consumer orders How the supply chain reacts
Change in order behaviour from 100 units to
400 units
Peak order of over 30,000 units!
SURPLUS IN A TYPICAL GAME
THE FEEDBACK LOOP GOVERNING THE SUPPLY CHAIN
THE FEEDBACK LOOP GOVERNING THE SUPPLY CHAIN
The problem: Each player can only control part of the control loop
THE ERROR MOST PEOPLE MAKE: INCLUDING THE BACK ORDER
IMPROVEMENT STRATEGY 1: IGNORE BACK ORDERS
Outgoing Orders = Incoming Orders + Target Inventory - Inventory
IMPROVEMENT STRATEGY 1: IGNORE BACK ORDERS
IMPROVEMENT STRATEGY 1: IGNORE BACK ORDERS
IMPROVEMENT STRATEGY 2: REMEMBER OPEN ORDERS
Remember the orders that are in the supply line
IMPROVEMENT STRATEGY 2: REMEMBER OPEN ORDERS
Remember the orders that are in the supply line
THE TARGET SUPPLY LINE DEPENDS ON THE DELIVERY DELAY
Target Supply Line = Delivery Delay * Incoming Orders
Orders = Incoming Orders + Target Inventory - Inventory + Target Supply Line - Supply Line
ORDER BEHAVIOUR WITH THE NEW ORDERING POLICY
THE INDIVIDUAL COST TARGET IS NOW REACHED
THE WHIPLASH EFFECT
Even if every player
behaves rationally,
there will still be a
"whiplash" effect –
because the orders
become successively
larger along the
supply chain
SUPPLY CHAIN COSTS ARE STILL WAY OFF TARGET
IMPROVEMENT STRATEGY 3: INCREASE INVENTORY ADJUSTMENT TIME
The solution to
dealing with the
whiplash effect is to
adjust inventory
slowly!
IMPROVED ORDER BEHAVOUR
TARGET SUPPLY CHAIN COSTS ARE MET
TARGET RETAILER COSTS ARE ALSO MET
SUMMARY OF THE COMPUTATIONAL MODELING APPROACH
Build a computational model to capture your understanding of the system
Use simulations to test the effect of different policies
Find policies that help you reach your targets
Controlling complex systems is hard and even harder when you only have par al control
We o en don't behave ra onally when overwhelmed with informa on and under pressure
Some mes it pays to "have faith in the system"
Small changes can have large e ects
Even if everybody op mises locally, this doesn't necessarily lead to a global op mum
Concrete learnings om the Beer Game
TRAINING AI TO PLAY THE BEERGAME
An approach using reinforcement learning
The idea: use autonomous agents to play the game and train them using reinforcement learning, a
machine learning technique.
Agents have
information about
their environment.
They perform
actions and receive
rewards (or
punishments) for
them.
They learn through
trial and error.
AGENTS FOR THE BEER GAME
The agents for the Beer Game are very simple
Q-LEARNING
A reinforcement learning technique
Each agent has a "Q-
Table", which
defines the
expected total
reward for each
state and each
possible action.
Agents start with an
"empty" q-table and
then fill it by
learning through
trial and error.
Always choose theac on with thehighest reward!
THE KEY: SETTING THE RIGHT REWARDS
Much like with us human beings, we need to set the right rewards
A reward for reaching the cost targets (and some milestones along the way)
"Game over" as soon as cost targets are missed (to avoid wasting time and memory)
THE RESULTS
Ten episodes 37,500 Episodes (almost
there)
50,000 Episodes
The AI agents can now
play the game
THE WINNER IS ...
The agents outperform our initial ordering strategy!
But: the agents are
optimised towards the
concrete game
situation - they would
fail in a more dynamic
stetting.
The ordering strategy
we developed initially
is robust in all ordering
situations.
SUMMARY
Reinforcement learning algorithms are quite easy to implement.
Finding the right reward policies is difficult: it is hard to avoid setting rewards to
narrow (much like in real life).
BUT: If you can define clear objectives and rewards, using reinforcement learning in
combination with simulations can be very useful for automating control systems.
NEXT MEETUP AFTER THE SUMMER BREAK
Building An Adaptive Enterprise on 24.9.2020
What characteristics define an adaptive enterprise?
Which capabilities and buidling blocks do you need to create an adaptive
enterprise?
How can you get there?
An adaptive enterprise is an enterprise that can flexibly adapt to a
changing economic environment.
Save The Date!
"Virtual Co ee Lounge"
for post-event discussions
DR. OLIVER GRASL
Oliver co-founded in 1997 and
has been managing partner ever since.
After reading mathematics and theoretical
physics at Cambridge University (MA
Cantab) and the University of Innsbruck he
later specialised in business engineering at
the University of St. Gallen (Executive MBA,
Dr. oec.).
Oliver’s personal mission is to help his
clients to explore, re-design and transform
their enterprises using his expertise in
enterprise architecture and enterprise
analytics.
transentis
oliver.grasl@transen s.com
+49 173 6546727
+49 30 800937050

More Related Content

Similar to Using Simulations And AI to Optimize Supply Chains

PD ESSENTIAL NEW_FINAL.pptx
PD ESSENTIAL NEW_FINAL.pptxPD ESSENTIAL NEW_FINAL.pptx
PD ESSENTIAL NEW_FINAL.pptx
khaledhassan
 
danmcclary-pspresentation-katieboyle-171030115522.pdf
danmcclary-pspresentation-katieboyle-171030115522.pdfdanmcclary-pspresentation-katieboyle-171030115522.pdf
danmcclary-pspresentation-katieboyle-171030115522.pdf
ssuser3ee399
 
Why Big and Small Data Is Important by Google's Product Manager
Why Big and Small Data Is Important by Google's Product ManagerWhy Big and Small Data Is Important by Google's Product Manager
Why Big and Small Data Is Important by Google's Product Manager
Product School
 
Coca cola new vending machine
Coca cola new vending machineCoca cola new vending machine
Coca cola new vending machineRavi Kumar Tiwary
 
Bid robot DSP EN
Bid robot DSP ENBid robot DSP EN
Bid robot DSP EN
DigitalBee
 
Two Methods for Modeling LTV with a Spreadsheet
Two Methods for Modeling LTV with a SpreadsheetTwo Methods for Modeling LTV with a Spreadsheet
Two Methods for Modeling LTV with a Spreadsheet
Eric Seufert
 
TFC DTU 22 introduction.pptx
TFC DTU 22 introduction.pptxTFC DTU 22 introduction.pptx
TFC DTU 22 introduction.pptx
ssuser3ee399
 
Retail Strategy Ric Anderson9.16.10
Retail Strategy Ric Anderson9.16.10Retail Strategy Ric Anderson9.16.10
Retail Strategy Ric Anderson9.16.10ricaaa1
 
[DSC Europe 23][Pandora] Siyu SUN Data Science Enter The Game.pptx
[DSC Europe 23][Pandora] Siyu SUN Data Science Enter The Game.pptx[DSC Europe 23][Pandora] Siyu SUN Data Science Enter The Game.pptx
[DSC Europe 23][Pandora] Siyu SUN Data Science Enter The Game.pptx
DataScienceConferenc1
 
Long Live Performance Marketing_Nick Cristal
Long Live Performance Marketing_Nick CristalLong Live Performance Marketing_Nick Cristal
Long Live Performance Marketing_Nick Cristal
PerformanceIN
 
Sunlight documentation
Sunlight documentationSunlight documentation
Sunlight documentation
Shehzad Yaqoob
 
Capstone PowerPoint
Capstone PowerPointCapstone PowerPoint
Capstone PowerPoint
Melis Jordan
 
Beer game esristhi
Beer game esristhiBeer game esristhi
Beer game esristhivijay kumar
 
Dirty Little Startup Secret: Barriers to entry
Dirty Little Startup Secret: Barriers to entryDirty Little Startup Secret: Barriers to entry
Dirty Little Startup Secret: Barriers to entry
Eric Tachibana
 
B2B Marketers: Sharpen Your Storytelling Skills (Workshop)
B2B Marketers: Sharpen Your Storytelling Skills (Workshop)B2B Marketers: Sharpen Your Storytelling Skills (Workshop)
B2B Marketers: Sharpen Your Storytelling Skills (Workshop)
LookBookHQ
 
Amzore Presentation : Dark Color Version
Amzore Presentation : Dark Color VersionAmzore Presentation : Dark Color Version
Amzore Presentation : Dark Color Version
Madlis
 
SAS Analytics2011 Retail final LI Post
SAS Analytics2011 Retail final LI PostSAS Analytics2011 Retail final LI Post
SAS Analytics2011 Retail final LI PostEmmett Cox
 
Amzore Presentation : Light Color Version
Amzore Presentation : Light Color VersionAmzore Presentation : Light Color Version
Amzore Presentation : Light Color Version
Madlis
 
Sales Forecasting for Management Consultants & Business Analysts
Sales Forecasting for Management Consultants & Business AnalystsSales Forecasting for Management Consultants & Business Analysts
Sales Forecasting for Management Consultants & Business Analysts
Asen Gyczew
 
Evaluation of the skillsmart retail Game
Evaluation of the skillsmart retail GameEvaluation of the skillsmart retail Game
Evaluation of the skillsmart retail Game
Lisa Eldred
 

Similar to Using Simulations And AI to Optimize Supply Chains (20)

PD ESSENTIAL NEW_FINAL.pptx
PD ESSENTIAL NEW_FINAL.pptxPD ESSENTIAL NEW_FINAL.pptx
PD ESSENTIAL NEW_FINAL.pptx
 
danmcclary-pspresentation-katieboyle-171030115522.pdf
danmcclary-pspresentation-katieboyle-171030115522.pdfdanmcclary-pspresentation-katieboyle-171030115522.pdf
danmcclary-pspresentation-katieboyle-171030115522.pdf
 
Why Big and Small Data Is Important by Google's Product Manager
Why Big and Small Data Is Important by Google's Product ManagerWhy Big and Small Data Is Important by Google's Product Manager
Why Big and Small Data Is Important by Google's Product Manager
 
Coca cola new vending machine
Coca cola new vending machineCoca cola new vending machine
Coca cola new vending machine
 
Bid robot DSP EN
Bid robot DSP ENBid robot DSP EN
Bid robot DSP EN
 
Two Methods for Modeling LTV with a Spreadsheet
Two Methods for Modeling LTV with a SpreadsheetTwo Methods for Modeling LTV with a Spreadsheet
Two Methods for Modeling LTV with a Spreadsheet
 
TFC DTU 22 introduction.pptx
TFC DTU 22 introduction.pptxTFC DTU 22 introduction.pptx
TFC DTU 22 introduction.pptx
 
Retail Strategy Ric Anderson9.16.10
Retail Strategy Ric Anderson9.16.10Retail Strategy Ric Anderson9.16.10
Retail Strategy Ric Anderson9.16.10
 
[DSC Europe 23][Pandora] Siyu SUN Data Science Enter The Game.pptx
[DSC Europe 23][Pandora] Siyu SUN Data Science Enter The Game.pptx[DSC Europe 23][Pandora] Siyu SUN Data Science Enter The Game.pptx
[DSC Europe 23][Pandora] Siyu SUN Data Science Enter The Game.pptx
 
Long Live Performance Marketing_Nick Cristal
Long Live Performance Marketing_Nick CristalLong Live Performance Marketing_Nick Cristal
Long Live Performance Marketing_Nick Cristal
 
Sunlight documentation
Sunlight documentationSunlight documentation
Sunlight documentation
 
Capstone PowerPoint
Capstone PowerPointCapstone PowerPoint
Capstone PowerPoint
 
Beer game esristhi
Beer game esristhiBeer game esristhi
Beer game esristhi
 
Dirty Little Startup Secret: Barriers to entry
Dirty Little Startup Secret: Barriers to entryDirty Little Startup Secret: Barriers to entry
Dirty Little Startup Secret: Barriers to entry
 
B2B Marketers: Sharpen Your Storytelling Skills (Workshop)
B2B Marketers: Sharpen Your Storytelling Skills (Workshop)B2B Marketers: Sharpen Your Storytelling Skills (Workshop)
B2B Marketers: Sharpen Your Storytelling Skills (Workshop)
 
Amzore Presentation : Dark Color Version
Amzore Presentation : Dark Color VersionAmzore Presentation : Dark Color Version
Amzore Presentation : Dark Color Version
 
SAS Analytics2011 Retail final LI Post
SAS Analytics2011 Retail final LI PostSAS Analytics2011 Retail final LI Post
SAS Analytics2011 Retail final LI Post
 
Amzore Presentation : Light Color Version
Amzore Presentation : Light Color VersionAmzore Presentation : Light Color Version
Amzore Presentation : Light Color Version
 
Sales Forecasting for Management Consultants & Business Analysts
Sales Forecasting for Management Consultants & Business AnalystsSales Forecasting for Management Consultants & Business Analysts
Sales Forecasting for Management Consultants & Business Analysts
 
Evaluation of the skillsmart retail Game
Evaluation of the skillsmart retail GameEvaluation of the skillsmart retail Game
Evaluation of the skillsmart retail Game
 

Recently uploaded

1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx
1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx
1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx
Tiktokethiodaily
 
The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...
jerlynmaetalle
 
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
AbhimanyuSinha9
 
Empowering Data Analytics Ecosystem.pptx
Empowering Data Analytics Ecosystem.pptxEmpowering Data Analytics Ecosystem.pptx
Empowering Data Analytics Ecosystem.pptx
benishzehra469
 
一比一原版(BU毕业证)波士顿大学毕业证成绩单
一比一原版(BU毕业证)波士顿大学毕业证成绩单一比一原版(BU毕业证)波士顿大学毕业证成绩单
一比一原版(BU毕业证)波士顿大学毕业证成绩单
ewymefz
 
Machine learning and optimization techniques for electrical drives.pptx
Machine learning and optimization techniques for electrical drives.pptxMachine learning and optimization techniques for electrical drives.pptx
Machine learning and optimization techniques for electrical drives.pptx
balafet
 
一比一原版(CU毕业证)卡尔顿大学毕业证成绩单
一比一原版(CU毕业证)卡尔顿大学毕业证成绩单一比一原版(CU毕业证)卡尔顿大学毕业证成绩单
一比一原版(CU毕业证)卡尔顿大学毕业证成绩单
yhkoc
 
一比一原版(TWU毕业证)西三一大学毕业证成绩单
一比一原版(TWU毕业证)西三一大学毕业证成绩单一比一原版(TWU毕业证)西三一大学毕业证成绩单
一比一原版(TWU毕业证)西三一大学毕业证成绩单
ocavb
 
Predicting Product Ad Campaign Performance: A Data Analysis Project Presentation
Predicting Product Ad Campaign Performance: A Data Analysis Project PresentationPredicting Product Ad Campaign Performance: A Data Analysis Project Presentation
Predicting Product Ad Campaign Performance: A Data Analysis Project Presentation
Boston Institute of Analytics
 
做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样
做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样
做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样
axoqas
 
一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单
一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单
一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单
ewymefz
 
Sample_Global Non-invasive Prenatal Testing (NIPT) Market, 2019-2030.pdf
Sample_Global Non-invasive Prenatal Testing (NIPT) Market, 2019-2030.pdfSample_Global Non-invasive Prenatal Testing (NIPT) Market, 2019-2030.pdf
Sample_Global Non-invasive Prenatal Testing (NIPT) Market, 2019-2030.pdf
Linda486226
 
standardisation of garbhpala offhgfffghh
standardisation of garbhpala offhgfffghhstandardisation of garbhpala offhgfffghh
standardisation of garbhpala offhgfffghh
ArpitMalhotra16
 
Criminal IP - Threat Hunting Webinar.pdf
Criminal IP - Threat Hunting Webinar.pdfCriminal IP - Threat Hunting Webinar.pdf
Criminal IP - Threat Hunting Webinar.pdf
Criminal IP
 
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...
pchutichetpong
 
一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单
一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单
一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单
ewymefz
 
一比一原版(UofS毕业证书)萨省大学毕业证如何办理
一比一原版(UofS毕业证书)萨省大学毕业证如何办理一比一原版(UofS毕业证书)萨省大学毕业证如何办理
一比一原版(UofS毕业证书)萨省大学毕业证如何办理
v3tuleee
 
一比一原版(QU毕业证)皇后大学毕业证成绩单
一比一原版(QU毕业证)皇后大学毕业证成绩单一比一原版(QU毕业证)皇后大学毕业证成绩单
一比一原版(QU毕业证)皇后大学毕业证成绩单
enxupq
 
一比一原版(YU毕业证)约克大学毕业证成绩单
一比一原版(YU毕业证)约克大学毕业证成绩单一比一原版(YU毕业证)约克大学毕业证成绩单
一比一原版(YU毕业证)约克大学毕业证成绩单
enxupq
 
FP Growth Algorithm and its Applications
FP Growth Algorithm and its ApplicationsFP Growth Algorithm and its Applications
FP Growth Algorithm and its Applications
MaleehaSheikh2
 

Recently uploaded (20)

1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx
1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx
1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx
 
The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...
 
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
 
Empowering Data Analytics Ecosystem.pptx
Empowering Data Analytics Ecosystem.pptxEmpowering Data Analytics Ecosystem.pptx
Empowering Data Analytics Ecosystem.pptx
 
一比一原版(BU毕业证)波士顿大学毕业证成绩单
一比一原版(BU毕业证)波士顿大学毕业证成绩单一比一原版(BU毕业证)波士顿大学毕业证成绩单
一比一原版(BU毕业证)波士顿大学毕业证成绩单
 
Machine learning and optimization techniques for electrical drives.pptx
Machine learning and optimization techniques for electrical drives.pptxMachine learning and optimization techniques for electrical drives.pptx
Machine learning and optimization techniques for electrical drives.pptx
 
一比一原版(CU毕业证)卡尔顿大学毕业证成绩单
一比一原版(CU毕业证)卡尔顿大学毕业证成绩单一比一原版(CU毕业证)卡尔顿大学毕业证成绩单
一比一原版(CU毕业证)卡尔顿大学毕业证成绩单
 
一比一原版(TWU毕业证)西三一大学毕业证成绩单
一比一原版(TWU毕业证)西三一大学毕业证成绩单一比一原版(TWU毕业证)西三一大学毕业证成绩单
一比一原版(TWU毕业证)西三一大学毕业证成绩单
 
Predicting Product Ad Campaign Performance: A Data Analysis Project Presentation
Predicting Product Ad Campaign Performance: A Data Analysis Project PresentationPredicting Product Ad Campaign Performance: A Data Analysis Project Presentation
Predicting Product Ad Campaign Performance: A Data Analysis Project Presentation
 
做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样
做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样
做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样
 
一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单
一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单
一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单
 
Sample_Global Non-invasive Prenatal Testing (NIPT) Market, 2019-2030.pdf
Sample_Global Non-invasive Prenatal Testing (NIPT) Market, 2019-2030.pdfSample_Global Non-invasive Prenatal Testing (NIPT) Market, 2019-2030.pdf
Sample_Global Non-invasive Prenatal Testing (NIPT) Market, 2019-2030.pdf
 
standardisation of garbhpala offhgfffghh
standardisation of garbhpala offhgfffghhstandardisation of garbhpala offhgfffghh
standardisation of garbhpala offhgfffghh
 
Criminal IP - Threat Hunting Webinar.pdf
Criminal IP - Threat Hunting Webinar.pdfCriminal IP - Threat Hunting Webinar.pdf
Criminal IP - Threat Hunting Webinar.pdf
 
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...
 
一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单
一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单
一比一原版(UofM毕业证)明尼苏达大学毕业证成绩单
 
一比一原版(UofS毕业证书)萨省大学毕业证如何办理
一比一原版(UofS毕业证书)萨省大学毕业证如何办理一比一原版(UofS毕业证书)萨省大学毕业证如何办理
一比一原版(UofS毕业证书)萨省大学毕业证如何办理
 
一比一原版(QU毕业证)皇后大学毕业证成绩单
一比一原版(QU毕业证)皇后大学毕业证成绩单一比一原版(QU毕业证)皇后大学毕业证成绩单
一比一原版(QU毕业证)皇后大学毕业证成绩单
 
一比一原版(YU毕业证)约克大学毕业证成绩单
一比一原版(YU毕业证)约克大学毕业证成绩单一比一原版(YU毕业证)约克大学毕业证成绩单
一比一原版(YU毕业证)约克大学毕业证成绩单
 
FP Growth Algorithm and its Applications
FP Growth Algorithm and its ApplicationsFP Growth Algorithm and its Applications
FP Growth Algorithm and its Applications
 

Using Simulations And AI to Optimize Supply Chains

  • 1. USING SIMULATIONS AND AI TO OPTIMIZE SUPPLY CHAINS Illustrated using the Beer Distribution Game Berlin, 28.5.2020 Transforming Enterprises
  • 2. TOPICS TODAY Introduc on to the Beer Distribu on Game, a supply chain simula on. Use computa onal modelling to understand supply chain dynamics and nd better playing strategies. Train autonomous agents to play the game using a machine learning / reinforcement learning approach
  • 3. DOWNLOAD THE CODE AND COMPUTATIONAL NOTEBOOKS Download all the code and in-depth computational notebooks from .GitHub All resources are available via www.transforming-enterprises.com
  • 4. THE BEER DISTRIBUTION GAME An Introduction The Beer Game was developed in the 1960s at MIT to illustrate how difficult it is to manage dynamic systems – in this case a supply chain that delivers beer from a brewery to the end consumer The game became well known in the 1990s after Peter Senge's description of it in his world-wide best selling book The Fifth Discipline The game is great because even though it is about a very simple system and despite very simple rules, the resulting behaviour is quite complex. – – –
  • 5. A BRIEF WALKTHROUGH OF THE BEER GAME Try it yourself at beergame.transentis.com The game is usually played with four players, but the single player version is also fun. But let's look at the rules and some poten al pi alls rst!
  • 6. THE SITUATION You are part of a supply chain that delivers beer from a brewery to the end consumer try to meet the demand of your respective customer at all times, while keeping inventory low. – –
  • 7. THE RULES The rules of the game are simple – in every round you perform the following steps: 1. Check deliveries. Check how many units of beer are being delivery to you from your supplier in the supply chain. 2. Check orders. Check how many units of beer your client in the supply chain has ordered. 3. Deliver beer. Deliver as much beer as you can to satisfy demand (the game does this for you). 4. Make an order decision. Decide how many units of beer you need from your supplier to keep your inventory stocked up. 1 2 3 4
  • 8. SOME PITFALLS TO BE AWARE OF Backorder and inventory costs Delays
  • 9. HOW PERFORMANCE IS APPRAISED Individual Supply Chain Cost. Your accumulated cost should remain below $8.300. Overall Supply Chain Costs. Total supply chain cost should remain below $29.300.
  • 10. LET'S GIVE IT A GO https://beergame.transentis.com
  • 11. ORDER BEHAVIOUR IN A TYPICAL GAME What the consumer orders How the supply chain reacts Change in order behaviour from 100 units to 400 units Peak order of over 30,000 units!
  • 12. SURPLUS IN A TYPICAL GAME
  • 13. THE FEEDBACK LOOP GOVERNING THE SUPPLY CHAIN
  • 14. THE FEEDBACK LOOP GOVERNING THE SUPPLY CHAIN The problem: Each player can only control part of the control loop
  • 15. THE ERROR MOST PEOPLE MAKE: INCLUDING THE BACK ORDER
  • 16. IMPROVEMENT STRATEGY 1: IGNORE BACK ORDERS Outgoing Orders = Incoming Orders + Target Inventory - Inventory
  • 17. IMPROVEMENT STRATEGY 1: IGNORE BACK ORDERS
  • 18. IMPROVEMENT STRATEGY 1: IGNORE BACK ORDERS
  • 19. IMPROVEMENT STRATEGY 2: REMEMBER OPEN ORDERS Remember the orders that are in the supply line
  • 20. IMPROVEMENT STRATEGY 2: REMEMBER OPEN ORDERS Remember the orders that are in the supply line
  • 21. THE TARGET SUPPLY LINE DEPENDS ON THE DELIVERY DELAY Target Supply Line = Delivery Delay * Incoming Orders Orders = Incoming Orders + Target Inventory - Inventory + Target Supply Line - Supply Line
  • 22. ORDER BEHAVIOUR WITH THE NEW ORDERING POLICY
  • 23. THE INDIVIDUAL COST TARGET IS NOW REACHED
  • 24. THE WHIPLASH EFFECT Even if every player behaves rationally, there will still be a "whiplash" effect – because the orders become successively larger along the supply chain
  • 25. SUPPLY CHAIN COSTS ARE STILL WAY OFF TARGET
  • 26. IMPROVEMENT STRATEGY 3: INCREASE INVENTORY ADJUSTMENT TIME The solution to dealing with the whiplash effect is to adjust inventory slowly!
  • 28. TARGET SUPPLY CHAIN COSTS ARE MET
  • 29. TARGET RETAILER COSTS ARE ALSO MET
  • 30. SUMMARY OF THE COMPUTATIONAL MODELING APPROACH Build a computational model to capture your understanding of the system Use simulations to test the effect of different policies Find policies that help you reach your targets Controlling complex systems is hard and even harder when you only have par al control We o en don't behave ra onally when overwhelmed with informa on and under pressure Some mes it pays to "have faith in the system" Small changes can have large e ects Even if everybody op mises locally, this doesn't necessarily lead to a global op mum Concrete learnings om the Beer Game
  • 31. TRAINING AI TO PLAY THE BEERGAME An approach using reinforcement learning The idea: use autonomous agents to play the game and train them using reinforcement learning, a machine learning technique. Agents have information about their environment. They perform actions and receive rewards (or punishments) for them. They learn through trial and error.
  • 32. AGENTS FOR THE BEER GAME The agents for the Beer Game are very simple
  • 33. Q-LEARNING A reinforcement learning technique Each agent has a "Q- Table", which defines the expected total reward for each state and each possible action. Agents start with an "empty" q-table and then fill it by learning through trial and error. Always choose theac on with thehighest reward!
  • 34. THE KEY: SETTING THE RIGHT REWARDS Much like with us human beings, we need to set the right rewards A reward for reaching the cost targets (and some milestones along the way) "Game over" as soon as cost targets are missed (to avoid wasting time and memory)
  • 35. THE RESULTS Ten episodes 37,500 Episodes (almost there) 50,000 Episodes The AI agents can now play the game
  • 36. THE WINNER IS ... The agents outperform our initial ordering strategy! But: the agents are optimised towards the concrete game situation - they would fail in a more dynamic stetting. The ordering strategy we developed initially is robust in all ordering situations.
  • 37. SUMMARY Reinforcement learning algorithms are quite easy to implement. Finding the right reward policies is difficult: it is hard to avoid setting rewards to narrow (much like in real life). BUT: If you can define clear objectives and rewards, using reinforcement learning in combination with simulations can be very useful for automating control systems.
  • 38. NEXT MEETUP AFTER THE SUMMER BREAK Building An Adaptive Enterprise on 24.9.2020 What characteristics define an adaptive enterprise? Which capabilities and buidling blocks do you need to create an adaptive enterprise? How can you get there? An adaptive enterprise is an enterprise that can flexibly adapt to a changing economic environment. Save The Date! "Virtual Co ee Lounge" for post-event discussions
  • 39. DR. OLIVER GRASL Oliver co-founded in 1997 and has been managing partner ever since. After reading mathematics and theoretical physics at Cambridge University (MA Cantab) and the University of Innsbruck he later specialised in business engineering at the University of St. Gallen (Executive MBA, Dr. oec.). Oliver’s personal mission is to help his clients to explore, re-design and transform their enterprises using his expertise in enterprise architecture and enterprise analytics. transentis oliver.grasl@transen s.com +49 173 6546727 +49 30 800937050