warehouse simulation: quick quick and effectiveand effective · 2019. 2. 14. · 2010 98.04% 77...

Warehouse Simulation: Warehouse Simulation:

Quick Quick and effectiveand effective

Alain de Norman et d`Audenhove – Director

Leandro Filippi – Consultant

BELGE - Sao Paulo / Brazil

About Belge

Objectives

Warehouse Environment & Simulation Technology

Agenda


Cases & Results

Technology behind & Final Comments

About Belge

Objectives

Warehouse Environment & Simulation TechnologyWarehouse Environment & Simulation Technology

Cases & Results


Belge Engenharia e Sistemas LTDA www.belge.com.br

Belge Engenharia e Sistemas LTDA www.belge.com.br

• Since 1995, Promodel Corp. Distributor for Mercosur(Brazil, Argentina, Paraguay and Uruguay)

• Consulting and software company specialized in

• Since 1995, Promodel Corp. Distributor for Mercosur(Brazil, Argentina, Paraguay and Uruguay)

• Consulting and software company specialized in

About Belge

• Consulting and software company specialized in quantitative methods

• Founded in 1995, originated from the central Engineering Department of Siemens

• Consulting and software company specialized in quantitative methods

• Founded in 1995, originated from the central Engineering Department of Siemens

Some Clients

About Belge

Introduction


Cases & Results


Warehouse environments include complexity, high levels of processinterdependency and variability. This makes the planning of distributioncenters an ideal environment for using dynamic simulation, though manylogistics operators continue to rely on static spreadsheets and theiroversimplifications.

WMS systems are excellent for managing operations, but are notsuitable for planning. They have no ability to test layouts, experimentwith different process alternatives, determine the right number of

Warehouse environments include complexity, high levels of processinterdependency and variability. This makes the planning of distributioncenters an ideal environment for using dynamic simulation, though manylogistics operators continue to rely on static spreadsheets and theiroversimplifications.

WMS systems are excellent for managing operations, but are notsuitable for planning. They have no ability to test layouts, experimentwith different process alternatives, determine the right number of

Introduction

with different process alternatives, determine the right number oftransport and human resources, or forecast the impact of differentdemand levels.

We show several ProModel simulations of DCs in large companies suchas Coca-Cola, Unilever, Brasil Foods and Colgate. Our focus is on theresults achieved for both new DC design and productivity improvementin existing facilities.

with different process alternatives, determine the right number oftransport and human resources, or forecast the impact of differentdemand levels.

We show several ProModel simulations of DCs in large companies suchas Coca-Cola, Unilever, Brasil Foods and Colgate. Our focus is on theresults achieved for both new DC design and productivity improvementin existing facilities.

About Belge

Introduction


Cases & Results


WarehouseWarehouseWarehouseWarehouse EnvironmentEnvironmentEnvironmentEnvironment & & & & SimulationSimulationSimulationSimulation TechnologyTechnologyTechnologyTechnology

Docks

Reception

Basic flow of a warehouse

Racks / Block

Stacking

Picking

Docks

Parking Area

SimulationSimulationSimulationSimulation considersconsidersconsidersconsiders thethethethe effecteffecteffecteffect ofofofof operationsoperationsoperationsoperations over a over a over a over a periodperiodperiodperiod ofofofof time!!!time!!!time!!!time!!!

How could a static spreadsheet (MS Excel) consider the fact that a forklift

can do several things at once if it does not simulate the operation over

time?

How could a static spreadsheet (MS Excel) consider the fact that a forklift

can do several things at once if it does not simulate the operation over

time?

How long does it take to unload a pallet from a rack?

Spreadsheets consider average time

Simulation allows us to consider lift time

AbilityAbilityAbilityAbility to to to to dealdealdealdeal withwithwithwith randomrandomrandomrandom variablesvariablesvariablesvariablesConsidersConsidersConsidersConsiders variabilityvariabilityvariabilityvariability on times and on times and on times and on times and speedsspeedsspeedsspeeds

Simulation allows us to consider lift time variability due to different numbers of levels per rack.

Static Analaysis: PI is input data

Static Analaysis: PI is input data

Dynamic analysis: PI is output data

Dynamic analysis: PI is output data

Eg.: New picking

system

The previous productivity

index was X boxes/hour

Assumes that with the new

The previous productivity

index was X boxes/hour

According to the new times,

resources qty. and the new

Productivity Index (PI)Productivity Index (PI)Productivity Index (PI)Productivity Index (PI)

system Assumes that with the newmethod, the productivity

will be increased to Y

boxes/hour

Scales the number of

operators and equipments

with the assumed

productivity

resources qty. and the new

method, the productivity index will

be Y boxes/hour

Evaluates the productivity (meets

or not) according on the amount

of resources and process times

SpreadsheetSpreadsheetSpreadsheetSpreadsheet staticstaticstaticstatic analysisanalysisanalysisanalysis vs.vs.vs.vs.SimulationSimulationSimulationSimulation dynamicdynamicdynamicdynamic analysisanalysisanalysisanalysis

Static analysis Dynamic analysis

Operations sequencing Does not consider Considered

Simultaneous request of

the same resource

Does not consider Considered

Variability in operative

times, speeds, demands

and breakdowns

Does not consider Considered

Productivity indices (eg

pallets/hour qty in

picking and load

assembly

It is an input data, wich

causes error. It is a

mistake to use these

values as a premise,

since it is influenced by

several factors such as:

resource qty, processing

times, arrivals and

departure sequence, etc.

It is an output of the

model. The result of all

its restrictions,

randomness, cycles,

demands, resources,

among others.


Warehouse enrivonments include:

Complexity

High level of process interdependency

Variability

Warehouse enrivonments include:

Complexity

High level of process interdependency

Variability

When to simulate?When to simulate?


What about WMS?

• Excellent for managing operations, but are not suitable forplanning.

• They have no ability to test layouts, experiment with differentprocess alternatives, determine the right number of transportand human resources or forecast the impact of different demandlevels.

What about WMS?

• Excellent for managing operations, but are not suitable forplanning.

• They have no ability to test layouts, experiment with differentprocess alternatives, determine the right number of transportand human resources or forecast the impact of different demandlevels.levels.levels.

WMS = managementWMS = management SIMULATION = planningSIMULATION = planning

About Belge

Objectives


Cases & Results


Some DCSim/ProMocel customers in South America

ObjectiveObjectiveObjectiveObjective::::

Audit 3 different operation logistics proposals and help Unilever to decidewhich 3PL would operate their new biggest DC in South AmericaJoint Warehouse (Foods and HPC / SP) – almost 60% of Unilever Renenue

(Brazil) comes from this DC

ObjectiveObjectiveObjectiveObjective::::

Audit 3 different operation logistics proposals and help Unilever to decidewhich 3PL would operate their new biggest DC in South AmericaJoint Warehouse (Foods and HPC / SP) – almost 60% of Unilever Renenue

(Brazil) comes from this DC

Case Unilever

Expansion - DHL Greenfield - DHL Greenfield - Mclane

Case - UnileverInteractive analysis to define the number ofresources required:

Inbound and outbound results X targets

Case Unilever

Eliminating

bottlenecks

and reducing

resources

idle

Outbound target: 5800 t

Inbound target: 4000 t

ErrorsErrors: : SizingSizing in Excel in Excel vsvs SimulationSimulation

Case Unilever

ResultsResultsResultsResults::::

Defined the mininum number ofresources necessary (we detectedoversizing and subsizing)

Costs were reduced by USUSUSUS$$$$ 130130130130,,,,000000000000 ////monthmonthmonthmonth (reducing operators and forklifts)

ResultsResultsResultsResults::::

Defined the mininum number ofresources necessary (we detectedoversizing and subsizing)

Costs were reduced by USUSUSUS$$$$ 130130130130,,,,000000000000 ////monthmonthmonthmonth (reducing operators and forklifts)

Case Unilever

Layout winner

Identified and eliminated bottlenecks.Otherwise the DC would have a backlog ofalmost 35% on peak days

Guarantee ability to attain the level ofservice required. Start-up operations ranvery well

Identified and eliminated bottlenecks.Otherwise the DC would have a backlog ofalmost 35% on peak days

Guarantee ability to attain the level ofservice required. Start-up operations ranvery well

Case - SadiaOBJECTIVE:OBJECTIVE:OBJECTIVE:OBJECTIVE:OBJECTIVE:OBJECTIVE:OBJECTIVE:OBJECTIVE:

Sizing the number of resourcesnecessaryCapacity planning

RESULTSRESULTSRESULTSRESULTSRESULTSRESULTSRESULTSRESULTSHeadcount reduced in 35%Based on improvements suggested by

DCSim, the client could increase the DC

Case Brasil Foods

DCSim, the client could increase the DC capacity by 20% without additional resources or investmentInvesment could be reduced by almost

30%Flexible model: the client can test new

scenarios any time, quickly and effectively

Case - SadiaOBJECTIVE:OBJECTIVE:OBJECTIVE:OBJECTIVE:OBJECTIVE:OBJECTIVE:OBJECTIVE:OBJECTIVE:

Capacity planningSizing head-countIdentify the best operation model

RESULTSRESULTSRESULTSRESULTSRESULTSRESULTSRESULTSRESULTSPossibility to reduce the number of forklifts

by 55%For future demand, it is not necessary to

Case Colgate

For future demand, it is not necessary toincrease the number of all resources.Increasing only the number of checkers (realbottleneck), they could increase capacity forthe next few yearsAvoid spending money by changing rack

positionsIdentified the DC capacity (how long this

DC can operate in current configuration) andwhen they should invest in a new DC or in anexpansion.

16 Bottlers in Brazil16 Bottlers in Brazil

Objective:

Development of simulation models to optimize the

layout, flows and storage of the new area of the DC in

Taguatinga do Sul of BRASAL.

The model contemplates the storage area expansion

and through the results analysis it was possible to

identify some operational restrictions in the system.

Example: Idleness level of each resource in each shift.

Objective:

Development of simulation models to optimize the

layout, flows and storage of the new area of the DC in

Taguatinga do Sul of BRASAL.

The model contemplates the storage area expansion

and through the results analysis it was possible to

identify some operational restrictions in the system.

Example: Idleness level of each resource in each shift.

Scope Scope

Case Coca-Cola - BRASAL

Scope

The model considered the following processes:

Storage Area;

Picking Area;

Loading/Unloading Tunnel;

Receiving and Expedition of Product.

Scope

The model considered the following processes:

Storage Area;

Picking Area;

Loading/Unloading Tunnel;

Receiving and Expedition of Product.

Previous LayoutPrevious Layout


Production

Lines

Storage area“Docks” – 16

loading points

Space idleness

Picking

StagingPicking area was not working

New Proposed LayoutNew Proposed Layout


Storage area

Production

Lines

Loading points

Picking

Replenishment

Picking

Staging area

(new concept)

Loading points

(8)


6 a.m.: all trucks must be loaded

Layout

proposed

Staging area

required

Baseline:

several trucks

loaded after the

deadline


Picking: Baseline Picking: New Layout1 2

New product positions

3Decrease of 1.5

min/pallet


Reduction: 12.5%

Example: Loading Time – Vehicle“Mercado” Baseline

Mean time �� 1.6 HR

Layout Proposed

Mean time �� 1.4 HR

- New staging area

- Defined the number of resources required

- Increased picking productivity

Production –

pallets produced

during the day

loading

day night


Pallets for

Picking

replenishmentReplenishment

Pallets

From

Picking

Results:

Through the simulation results, bottlenecks were

identified, as operation problems and the breakpoint

of the DC in the current situation;

Based on the simulation results, a new layout was

proposed, improving storage capacity by 20%;

Results:

Through the simulation results, bottlenecks were

identified, as operation problems and the breakpoint

of the DC in the current situation;

Based on the simulation results, a new layout was

proposed, improving storage capacity by 20%;


The operation strategy of the DC was changed,

reducing vehicle loading times by almost 26%;

A new picking configuration was proposed, reducing

the picking time and inspection in the stages.

The operation strategy of the DC was changed,

reducing vehicle loading times by almost 26%;

A new picking configuration was proposed, reducing

the picking time and inspection in the stages.

Goal:

Development of simulation models that allow the

optimization the stock area of Porto Alegre Unit.

Determine the best layout considering some enlargement

area possibilities, identifying which layout best

accommodates growing demand until 2018.

Goal:

Development of simulation models that allow the

optimization the stock area of Porto Alegre Unit.

Determine the best layout considering some enlargement

area possibilities, identifying which layout best

accommodates growing demand until 2018.

Case Coke 2

Comparision of the proposed scenarios

% Customer Service Level

40m

Ideal

Summer

Tunnel

30m

Tunnel

30m

Lateral

Dock

60m

Frontal

Dock

60m

Lateral

Dock

2010 98.04% 77 97.83% 77 97.97% 77 99.00% 78 97.77% 77

2011 96.07% 81 97.95% 83 99.00% 84 96.02% 81

2012 91.92% 84 95.93% 87 98.00% 89 96.03% 87

2013 90.13% 89 93.78% 91 90.23% 89 94.01% 92

2014 87.06%* 93 - 88.85%* 95 84.87% 90 94.22%** 101

Results:

Case Coke 2

* with 18 forklifts (10

simple)

** adjusting resale window and route

service time

2014 87.06%* 93 - 88.85%* 95 84.87% 90 94.22%** 101

2015 - - 90.22%** 105

2016 - - 76.46%**

Goal:

Improve the Distribution Center capacity in order to

optimize stock areas and flows, considering the

business dynamics, seeking to decrease the

operational costs and improve the asset’s applications.

Simulation Period:

Goal:

Improve the Distribution Center capacity in order to

optimize stock areas and flows, considering the

business dynamics, seeking to decrease the

operational costs and improve the asset’s applications.

Simulation Period:

Case Coke 3

Simulation Period:

2009 to 2012;

Scope:

Storage areas, DC internal flows, picking area,

loading/unloading docks, DC Gates, factory output.

Simulation Period:

2009 to 2012;

Scope:

Storage areas, DC internal flows, picking area,

loading/unloading docks, DC Gates, factory output.

Results:

--18% Reduction of 18% Reduction of forklift forklift displacement distance (considering the same demand)

14% Reduction by: better positioning of products.

4% Reduction by: loading/unloading by forklifts’ circuit.

Average distance by Loaded Pallet (before simulation): 178m

Average distance by Loaded Pallet (after simulation): 153m

+8,9% +8,9% Increase in storage capacity

Case Coke 3

+8,9% +8,9% Increase in storage capacity

Previous Capacity : 15.833 pallets

Current Capacity : 17.233 pallets

--12% 12% Decrease of distributor loading time

Previous average time: 1:32 hours

Later average time: 1:21 hours

Objetivos Resultados

Optmize flows and

displacements

Increase the storage area utilization by

8.9%

Space optmization

Decrease of 18% in vehicls service

times

Size the number of resources

required for the next years Correct sizing of the resources required

Objectives Results

Some Results at Coke

Define the best layouts for

future scenarios

Required Investment reduced by

USD 9 million

Define the best operation

strategy

Best ‘summer plan’ of Coke/Vonpar

history

Set the best increasing strategy High precision and quality

Assis Brasil


Set the best layout for future and

increasing demands Obtained the optmized layout for the DC

Set the best operation strategy Charging system dramatically modified

Increase of 20% in warehouse area

Objectives Results

ANC


Set the best layout for increasing volumes

Increase of 20% in warehouse area

availability

Define the picking and stagin area Truck loading time reduced by 40%

Improvement in the picking area

Definition of a new layout for picking

area


Define the best layout for a

new DC

Defined warehouse area, restaurant,

parking area, etc

Identify the investiments

required in the future years Investiment plan per year

Sizing the number of resources

Sized the mininum number of human

resources and equipments

Increase of 45% in warehousing capacity

Objectives Results


Define the best layout for a

new DC

Increase of 45% in warehousing capacity

when compared with the original

proposed layout

Flow optimization

Defined the picking method, type of

storage structures, staging areas and set

the best regions for each SKU

Size the number of resouces Defined the minumun qty of resouces

“We could, thanks to the project made by Belge

in Vonpar, choose the best scenario considering a

Layout Definition and Optimizing the flows. The

result was a great increase in the operation level of

“We could, thanks to the project made by Belge

in Vonpar, choose the best scenario considering a

Layout Definition and Optimizing the flows. The


“A unique learning opportunity, where we

evolved a lot in the knowledge of our inner

processes, and mainly to know which results can be

pursued and leveraged.”

“A unique learning opportunity, where we

evolved a lot in the knowledge of our inner

processes, and mainly to know which results can be


Testimony


our DC concerning storage, displacement and service

level to our clients.”

“We can say that it was the best summer plan

in the history of Vonpar."

Sandro Soares

Logistics Coordinator


our DC concerning storage, displacement and service

level to our clients.”

“We can say that it was the best summer plan

in the history of Vonpar."

Sandro Soares

Logistics Coordinator


Heitor Ferreira Perez Villar

Logístics Analyst


Heitor Ferreira Perez Villar

Logístics Analyst

About Belge

Introduction


Cases & Results


• Developed as an in-house tool for quickly modeling DCs

• Dramatic reduction in time required

• Customer requests to continue using the tool on their own

What is DCSim

DCSim uses DCSim uses ProModelProModel Technology Technology

• World state-of-art in simulation

• Clear for all users (does not requires any previous knowledge about simulation)

Simulator

Language

Routing

DCSim Model

Subroutines• Inbound• Outbound• Picking• Conference• Storage area• etc

Input Data:• Resources• Times• Demand• Shifts• etc

How it works?

44

Cockpit –Input Datas

Results

Conclusions

Simulation is helping several companies to avoid typycal sizing errors (compared to simple MSExcel usage)

According to our experience in several DC projects, we can point to somesignificant improvements, such as:

• Up to a 40% increase in outbound capacity, made possible through theidentification and modification of system bottlenecks

• Minimized startup errors in new and modified DCs

• Up to a 30% reduction in human resource requirements during peak

Conclusions

days

• Operational cost reductions of up to 35%

ThankThankThankThank YouYouYouYou !!!!!!!!!!!!

QuestionsQuestionsQuestionsQuestions ????

Alain de Norman et d´Audenhove - [email protected]

Leandro Filippi – [email protected] Filippi – [email protected]

+55 (11) 5561-5353

www.belge.com.br

warehouse simulation: quick quick and effectiveand effective · 2019. 2. 14. · 2010 98.04% 77...

Documents