Trend in Supply Chain Optimization and Humanitarian Logistics Tokyo University of Tokyo University of Marine Science and Technology Marine Science and.

Trend in Supply Chain Optimization and Humanitarian Logistics

Tokyo University of Tokyo University of

Marine Science and Technology Marine Science and Technology

KUBO KUBO MikioMikio

Agenda

Definition of the Supply Chain (SC) and Logistics Decision Levels of the SC Classification of Inventory Basic Models in the SC Logistics Network Design Inventory Production Planning Vehicle Routing

SC Risk Management and Humanitarian SC

What’s the Supply Chain?

IT(Information Technology)＋ Logistics ＝ Supply Chain

Real System, Transactional IT, Analytic IT

実 シ ス テ ム

IT処 理 的

IT解 析 的

Real System=Truck, Ship, Plant, Product, Machine, …

Transactional ITPOS, ERP, MRP, DRP…Automatic Information Flow

Analytic ITModel ＋ Algorithm=Decision Support System

brain

muscle

nerve

Levels of Decision Making

Strategic LevelStrategic Level

Tactical LevelTactical Level

Operational LevelOperational Level

A year to several years; long-term decision making

A week to several months; mid-term decision making

Real time to several days; short-term decision making

Transactional IT

Analytic IT

Models in Analytic IT

Logistics Network Design

Inventory

Safety stock allocationInventory policy

optimization

Production

Lot-sizingScheduling

TransportationDelivery

Vehicle Routing

Multi-period Logistics Network Design

Strategic

Tactical

Operational

Plant DCSupplier Retailer

Models in Analytic IT

Logistics Network Design

Inventory

Safety stock allocationInventory policy

optimization

Production

Lot-sizingScheduling

TransportationDelivery

Vehicle Routing

Multi-period Logistics Network Design

Strategic

Tactical

Operational

Plant DCSupplier Retailer

Models in Analytic IT

Logistics Network Design

Inventory

Safety stock allocationInventory policy

optimization

Production

Lot-sizingScheduling

TransportationDelivery

Vehicle Routing

Multi-period Logistics Network Design

Strategic

Tactical

Operational

Plant DCSupplier Retailer

Inventory=Blood of Supply Chain

Time

Inventory acts as glue connecting optimization systems

Plant DCSupplier Retailer

Raw material Work-in-process Finished goods

Classification of Inventory

In-transit (pipeline) inventoryTrade-off: transportation cost or production speed

Seasonal inventoryTrade-off: resource acquisition or overtime cost ， setup cost

Cycle inventoryTrade-off : transportation (or production or ordering) fixed cost

Lot-size inventoryTrade-off: fixed cost

Safety inventoryTrade-off: customer service level ， backorder (stock-out) cost

In-transit (pipeline) Inventory

Inventory that are in-transit of products

Trade-off: transportation cost or transportation/production speed

->optimized in Logistics Network Design (LND)

Seasonal Inventory

Inventory for time-varying (seasonal) demands

Trade-off: resource acquisition or overtime cost -> optimized in multi-period LND

Trade-off: setup cost -> optimized in Lot-sizing

Demand

Period

Resource Upper Bound

Cycle InventoryInventory caused by periodic activities

Trade-off : transportation fixed cost -> LND

Trade-off: ordering fixed cost-> Economic Ordering Quantity (EOQ)Inventory

Level demand

Cycle Time

Lot-size Inventory

Cycle inventory when the speed of demand is not constant

Trade-off: fixed cost ->Lot-sizing, multi-period LND

Time

Inventory Level

Safety Inventory

Inventory for the demand variability

Trade-off: customer service level ->Safety Stock Allocation, LND

Trade-off: backorder (stock-out) cost ->Inventory Policy Optimization

Classification of Inventory

Time

Safety Inventory

Cycle InventoryLot-size Inventory

In-transit (Pipeline) Inventory

It’s hard to separate them but…They should be determined separately to optimize the trade-offs

Seasonal Inventory

Logistics Network DesignDecision support in strategic levelTotal optimization of overall supply chains

Example Where should we replenish pars? In which plant or on which production

line should we produce products? Where and by which transportation-

mode should we transport products? Where should we construct (or close)

plants or new distribution centers?

Trade-off in LND Model: Ｎ umber of Warehouses v.s.

輸 送 中 在 庫 費 用 輸 送 費 用Number of warehouses

• Service lead time ↓• Inventory cost ↑• Overhead cost ↑• Outbound transportation cost ↓• Inbound transportation cost ↑

Trade-off: In-transit inventory cost v.s. Transportation cost

輸 送 中 在 庫 費 用 輸 送 費 用In-transit inventory costTransportation cost

Multi-period Logistics Network Design

Decision support in tactical levelAn extension of MPS (Master Production System) for production to the Supply ChainTreat the seasonal demand explicitly

Demand

Period (Month)

Trade-off:Overtime v.s. Seasonal Inventory Cost

資 源 超 過 ペ ナ ル テ ィ（ 残 業 費 ）

作 り 置 き 在 庫 費 用Demand

Period

Resource Upper Bound

Constant Production

Inventories

VariableProduction

Overtime

Overtime penaltySeasonal inventory

S uppliersP lants

P roduc t ion L ines

Warehouses C ustomer G ropus

× ３

B O M or R ec ipie

Mixed Integer Programming (MIP) + Concave Cost Minimization

Safety Inv. CostBOM or Recipe

Safety Stock Allocation

Decision support in tactical levelDetermine the allocation of safety stocks in the SC for given service levels安 全 在 庫 費 用 サ ー ビ ス レ ベ ル

＋ 統 計 的 規 模 の 経 済（ リ ス ク 共 同 管 理 ）

Safety Inventory Service Level

+Risk Pooling (Statistical Economy of Scale)

Basic Principle of Inventory

Economy of scale in statistics: gathering inventory together reduces the total inventory volume. 　

-> Modern supply chain strategies risk pooling delayed differentiation design for logistics

Where should we allocate safety stocks to minimize the total safety stock costs so that the customer service levelis satisfied.

Lead-time and Safety Stock

Normal distribution with average demand μ ， standard deviation σService level （ the probability of no stocking out ） 95%->safety stock ratio 1.65

Lead-time (the time between order and arrival ） L

LL RatioStock Safety

Volume Inv.Max

＋

＝

The Relation between Lead-time and (Average, Safety, Maximum) Inventory

0

500

1000

1500

2000

2500

3000

0 5 10 15 20

Lead- time

AverageMax.Safety

Guaranteed Lead-timeGuaranteed lead-time (LT) ： Each facility guarantees to deliver the item to his customer within the guaranteed lead-time

Facility i

1 Production time Ti ＝ 3

2Guaranteed LTof upstream facility=1 day= Entering LT LIi

Guaranteed LT to downstream facility

Li ＝２ days

2

Safety inv.＝２ days

Net Replenishment TimeNet replenishment time (NRT) ：

　＝ LTi +Ti -Li

Facility i

1 Production time Ti ＝ 3

2Guaranteed LTof upstream facility=1 day= Entering LT LIi

2

Safety inv.＝２ days

Guaranteed LT to downstream facility

Li ＝２ days

Safety Stock Allocation Formulation

net replenishment time

maximum demand

upper bound of guaranteed LT

Algorithms for Safety Stock Allocation

Concave cost minimization using piece-wise linear approximationDynamic programming (DP) for tree networksMetaheuristicsLocal Search (LS), Iterated LS, Tabu Search

A Real Example: Ref. Managing the Supply Chain –The Definitive Guide for the Business Professional –by Simchi-Levi, Kaminski,Simchi-Levi

458 29

Part7Denver ($2.5)

Part 6Raleigh ($3)

58 837

37

37

3

28

37 171539

15

15 1530

30

Final DemandN(100,10)Guaranteed LT=30 days

5

Part 5Charleston ($12)

Part 4Malaysia ($180)

Part 3Montgomery ($220)

Part 2Dallas ($0.5)

Part 1Dallas ($260)

x2

43,508$ (40%Down)

What if analysis: Guaranteed LT=15 days ->51,136$

Inventory Policy Optimization

Decision support in operational/tactical levelDetermine various parameters for inventory control policies

品 切 れ 費 用 安 全 在 庫 費 用 発 注 （ 生 産 ） 固 定 費 用サ イ ク ル 在 庫 費 用

Classical Newsboy Model Classical Economic Ordering Quantity Model

Safety Inventory Lost Sales Cycle InventoryFixed Ordering

Base stock Policy (Multi Period Model)

Base stock level s* ＝ target of the inventory positionInventory (ordering) position=In-hand inventory+In-transit inventory (inventory on order) -BackorderBase stock policy: Monitoring the inventory position in real time; if it is below the base stock level, order the amount so that it recovers the base stock level

(Q,R) and (s,S) Policies

If the fixed ordering cost is positive, the ordering frequency must be considered explicitly.(Q,R) policy ： If the inventory position is below a re-ordering point R, order a fixed quantity Q (s,S) policy ： If the inventory position is below a re-ordering point s, order the amount so that it becomes an order-up-to level S

(Q,R) Policy and (s,S) Policy

R(=s)

R+Q(=S)

Lead time

(Q,R)

(s,S)Inventory position

Time

In-hand inventory

Periodic Ordering PolicyCheck the inventory position periodically; if it is below the base stock level, order the amount so that it recovers the base stock level

Mon. Tue. Wed. Thu.

Demand

Arrival of the order of Mon. （ Lead time L ＝１ day ）

Order

L=1

Algorithms for Inv. Policy Opt.

base stock ， (Q,R) ， and (s,S) policies->Dynamic Programming RecursionPeriodic ordering policy-> Infinitesimal Perturbation Analysis During simulation runs, derivatives of the cost function are estimated and are used in non-linear optimization

Lot-size OptimizationDecision support in tactical levelOptimize the trade-off between set-up cost and lot-size inventory

段 取 り 費 用 在 庫 費 用Lot-size Inv. Setup Cost

Algorithms for Lot-sizing

MIP solver with strong forumulation(Meta)heuristics Metaheuristics using MIP solver Relax and Fix Capacity scaling ＭＩＰ based neighborhood local search

Scheduling Optimization

Decision support in operational levelOptimization of the allocation of activities (jobs, tasks) over time under finite resources (such as machines)

Machine 1

Machine 2

Machine 3

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Time

What is Scheduling?

Allocation of activities (jobs, tasks) over time Resource constraints. For example, machines,

workers, raw material, etc. may be scare resources. 　

Precedence relation. For example., some activities cannot start unless other activities finish.

Machine 1

Machine 2

Machine 3

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Time

Solution Methods for Scheduling

Myopic heuristics Active schedule generation scheme Non-delay schedule generation scheme Dispatching rules

Constraint programming

Metaheuristics

Vehicle Routing Optimization

DepotDepot

CustomersCustomers

RoutesRoutes

service timewaitingtime

earliest time

Customer

latest time

service time

Algorithms for Vehicle Routing

Saving (Clarke-Wright) method Sweep (Gillet-Miller) method Insertion method Local Search Metaheuristics

History of Algorithms for Vehicle Routing Problem History of Algorithms for Vehicle Routing Problem

Construction MethodConstruction Method(Saving, Insertion)(Saving, Insertion)

Local SearchLocal Search

GeneralizedGeneralizedAssignmentAssignment

Location Based Location Based HeuristicsHeuristics

Tabu SearchTabu Search

GRASPGRASP(Greedy Randomized(Greedy RandomizedAdaptive Search Procedure)Adaptive Search Procedure)

AMPAMP(Adaptive Memory(Adaptive Memory Programming)Programming)

Approximate AlgorithmApproximate Algorithm

Exact AlgorithmExact Algorithm

K-Tree Relax.K-Tree Relax.

State Space Relax.State Space Relax.Set Partitioning ApproachSet Partitioning Approach

HierarchicalHierarchicalBuilding BlockBuilding BlockMethodMethod

Cutting PlaneCutting Plane

19701970 19801980 19901990

Route SelectionRoute SelectionHeuristicsHeuristics

Simulated AnnealingSimulated Annealing

Genetic AlgorithmGenetic Algorithm

20002000

SweepSweepMethodMethod

Disruption

Supply Chain “Risk” Management

Proactive and response approaches to cope with supply chain disruptions.

Time

Per

form

ance

Proactive Response

Recovery

Importance of Supply Chain “Risk”

Increase of disasters Natural disasters: earthquake, tsunami, SARS (Severe Ac

ute Respiratory Syndrome), BSE (Bovine Spongiform Encephalopathy), hurricanes, cyclones and typhoons, floods, droughts, volcanic eruption, famine and food insecurity, etc.

Man-made disasters: terrorist attack, CBRNE (Chemical Biological, Radiological, Nuclear, Explosive) disaster, war, strike, riot, etc.

Lean supply chain: increases vulnerability. Globalization: induces long lead time, outsourcing.

Related Area

Risk ManagementBusiness Continuity Planning (BCP)/ Business Continuity Management (BCM)

But, both did not work well …

Humanitarian Logistics / Humanitarian Supply Chain

Humanitarian Logistics / Humanitarian Supply Chain… is a branch of logistics which specializes in organiz

ing the delivery and warehousing of supplies during natural disasters to the affected area and people.DecentralizedNo SCM unit nor trained staffsEverything is ad hocNo performance measure (fairness, speed, …)No information & communication technologyMany players (government, NGOs)

Risk MappingRegular risk : demand/supply uncertainty Irregular risk : disruption / disaster

Impact

Frequency

Typhoon

Earthquake

Strike

Exchange Rate

Line Stop

Supply Delay

Defective Product

Risk Classification (1)

Supply Risk

Plant

ProductionLine

TransportationResource

Warehouse

Demand Risk

Internal Risk

Environmental Risk

Risk Classification (2)

Disaster risk: natural and man-made disasters such as landslides, volcanic eruption, drought, asteroid impactsPolitical risk: contracts, laws, regulations Social risk: child labor / abuse Intellectual property risk: patents, trademarks, copyrights Financial risk, employment risk, reputation risk, ...

Strategies to Cope with Risk

Accept: just do nothing!Avoid: remove the risk factor, if possibleTransfer: insurance, option Alignment: share risk and profit by contractStrengthen: make the SC robust, resilient, redundant, flexible, …

Strengthen StrategiesProactive Robustness Resiliency Redundancy Flexibility Compatibility

Disruption

Time

Per

form

ance

Proactive Response

• Response

– Agility

– Visibility

Robustness

Time Resiliency

PerformanceResiliency

Redundancy -Strategic Inventory-

Inventory for supply (or production) disruptions.That is shared by many supply chain partners.We have to distinguish it with the safety stock to copy with demand uncertainty.

Flexibility of Sourcing-Multiple Sourcing Strategy-

PlantSupplierSingle sourcing

Dual sourcing

Make-and-buy

Plant

Supplier A

Supplier B(Contract)

SupplierPlant

or

Flexible Production Strategy

Graves-Tomlin: 2-flex. is enough for demand uncertainty, i.e., 2-flex. has the similar performance with full-flex.Simulation : 2-flex. is NOT enough for supply uncertainty.

1-flexibility 2-flexibility Full-flexibility

Flexible Transportation Strategy

Multi-mode

Multi-carrier

Multi-route

Compatibility

Risk PoolingDelayed Differentiation / Postponement

Coping Strategies / Risk Mapping

Impact

Frequency

Typhoon

Earthquake

Strike

Exchange Rate

Line Stop

Supply Delay

Defective Product

Avoid

Transfer

Redundancy

FlexibilityRobustness by KAIZEN/TQC

VisibilityAlignment

Robustness by PM

Reduce Impact

Reduce P

robability

Supply Chain “Risk” Optimization

What If AnalysisStochastic Programming (Scenario Approach)

Disruption

Time

Per

form

ance

Proactive Response

Logistics Network DesignSafety Stock Allocation(Strategic, Tactical)

Scheduling Vehicle Routing Transportation(Operational)

Here & NowVariables Recourse Variables

Optimization Models for SCRM

Logistics Network Design

Inventory

Safety stock allocationInventory policy

optimization

Production

Lot-sizingScheduling

TransportationDelivery

Vehicle Routing

Multi-period Logistics Network Design

Strategic

Tactical

Operational

Stochastic /RobustExtensionsDynamic Pricing

Sourcing Decision

Quick Solution without IT