Outline

1
Outline

• ideas of benchmarking
• DEA
• profiling

2
Purpose of the Course

• warehouses and warehousing means, not ends
• ends for students
• satisfy the course requirement
• prepare for thesis
• how to collect information, present, write an
  essay
• self-improve and self-actualize

3
Thesis

• a serious issue
• certainly not something from cutting and pasting
• not merely a collection of organized material
• a step on generating knowledge
• material read serving as the basis
• key your own thoughts
• hard, but worthwhile training

4
Term Project

• the training for your thesis
• just try your best, and dont worry that much

5
Benchmarking and Profiling
6
Tasks for Senior Management of Warehouses

• continuous improvement
• setting objectives
• absolute standard, e.g., 95 orders in 2 days, on
  average no more than 2.2 days
• relative standard benchmarking
• profiling pre-requisite of benchmarking
• soul searching

7
Steps for Benchmarking

• identify the process to benchmark for e.g., most
  troublesome, most important
• identify the key performance variables
  efficiency (time, cost, productivity) and service
  level
• document current processes and flows physical
  activities and information flows
• including resources required
• identify competitors and best-in-class companies
• decide which practices to adopt
• see modifications

8
Data Collected for Benchmarking Warehouses

• performance benchmarking
• inputs, e.g.,
• labor, investment, space, scale of storage,
  degree of automation
• outputs
• of lines picked, level of value added service,
  of special processes, quality of service,
  flexibility of service
• broken case lines shipped, full case lines
  shipped and pallet lines shipped
• process benchmarking
• resources
• procedure
• results

9
Difficulties of Benchmarking

• intangible factors
• how to measure factors such as degree of
  automation, level of value added service, quality
  of service, flexibility of service, etc.
• incomparable factors
• e.g., the comparison of quality of service with
  degree of automation

10
Common Approaches for Intangible Factors

• qualitative description, e.g.,
• different levels of sophistication of receiving

Stage 1 measure Stage 3 Stage 4 Stage 5
Receiving unload, stage, in-check immediate putaway to reserve immediate putaway to primary cross-docking prereceiving
11
Steps to World-Class Warehousing Practices
12
Common Approaches for Intangible Factors

• numerical values assigned to qualitative factors
• quantitative measures for qualitative factors
• e.g., quality of service by of customers
  satisfied in 5 minutes, level of value added
  service by types of value added service provided

13
Examples of Numerical Performance Indicators
Based on Table 3-4 Warehouse Key Performance
Indicators (Frazell (2002))
Financial Productivity Utilization Quality Cycle time
Receiving
Putaway
Storage
Order picking
Shipping
Total
14
Examples of Numerical Performance Indicators
Based on Table 3-4 Warehouse Key Performance
Indicators (Frazell (2002))
Financial Productivity Utilization Quality Cycle time
Receiving Cost / line Receipts / man-hr Dock utilization of correct receipts processing time / receipt
Putaway Cost /line Putaway / man-hr Labor equipment utilization of perfect putaway Cycle time / putaway
Storage Cost / item Inv / area Space utilization of accurate record Inv. day
Order picking Cost / line Line picked / man-hr Labor equipment utilization of correct picked lines Pick cycle time
Shipping Cost / order Order shipped / man-hr Dock utilization of perfect shipments cycle time / order
Total Cost / order, line, item Lines shipped / man-hr --- of perfect W/H orders Cycle time / order
15
PresentingIncomparable Factors

• skipping comparison, e.g., the web graph for gap
  analysis
• an example for 6 factors
• best practices identified for benchmarking
• the relative performance with respect to the best
  praes

16
ComparingIncomparable Factors

• various methods, e.g., Scoring, Analytic
  Hierarchy Process, Balanced Scorecard, Data
  Envelopment Analysis (DEA), etc.

17
Data Envelopment Analysis (DEA)
18
Comparing Incomparable Factors

• data envelopment analysis (DEA) a technique to
  compare quantitative factors of different nature
• providing a numerical value judging the distance
  from the best practices
• some assumptions
• numerical values of each factor, e.g., input1
  5, input2 12, though input1 and input2 cannot
  be compared
• linearity of effect, i.e., if 3 units of input
  give 7 units of outputs, 6 units of input give 14
  units of output

19
Idea of Data Envelopment Analysis (DEA)

• W/H A and W/H B consume the same amount of
  resources
• two types of incomparable outputs apple and
  orange
• which is better?

20
Idea of Data Envelopment Analysis (DEA)

• W/H C consumes the same amount of resources as
  W/Hs A and B do
• Hows the performance of C relative to A and B?

21
Idea of Data Envelopment Analysis (DEA)

• Given W/H A and B, for W/Hs that consumes the
  same amount of resources, the inefficient region
  is shown in RHS.
• The efficiency of a warehouse that consumes the
  same amount of resources as A and B can be
  measured by the distance from the boundary of the
  date envelope.

22
Idea of Data Envelopment Analysis (DEA)

• efficient boundary from many warehouses that
  consume the same amount of resources

inefficient region
23
Idea of Data Envelopment Analysis (DEA)

• efficient boundary from many warehouses that give
  the same amount of outputs and consume different
  values of incomparable resources banana and
  grapefruit

inefficient region
24
Idea of Data Envelopment Analysis (DEA)

• problem situations for benchmarking often not
  ideal
• different resources consumption for W/H
• different outputs for W/H
• for multi-input, multi-output problems, with W/H
  consuming different amount of resources and
  giving different amount of outputs, DEA
• draws the efficient boundary
• benchmarks a W/H with respect to these existing
  ones

25
Idea of Data Envelopment Analysis (DEA)

• multi-input, multi-output comparison
• I decision-making units (DMUs), J types of
  inputs, K types of outputs
• aij be the number of units of input j that entity
  i takes to give aik units of output k, j 1, ,
  J and k J1, , JK
• example 2 DMUs 2 types of inputs (grapefruit,
  banana) 2 types of outputs (apple, orange)
• DMU 1 a11 1, a12 3, a13 5, and a14 2,
  i.e., DMU 1 takes 1 grapefruit, 3 bananas to
  produce 5 apples and 2 oranges
• DMU 2 a21 2, a22 1, a23 3, and a24 4,
  i.e., DMU 2 takes 2 grapefruits, 1 banana to
  produce 3 apples and 4 oranges

26
Idea of Data Envelopment Analysis (DEA)

• rk unit reward of type k output, cj unit cost
  of type j input
• performance of DMU 1 (5r32r4)/(c13c2)
• performance of DMU 2 (3r34r4)/(2c1c2)
• performance of DMU i defined similarly
• given (aij) of the I DMUs, how to benchmark a
  tapped DMU with (aoj) for unknown rk and cj?

27
Idea of Data Envelopment Analysis (DEA)

• in general DEA finds the distance from the
  efficient boundary by a linear program purely
  making use of (aij) and (aoj) without knowing rk,
  nor cj
• idea similar to the construction of efficient
  boundaries in the simplified examples

28
Studies Using DEA on Warehouses

• de Koster, M.B.M., and B.M. Balk (2008)
  Benchmarking and Monitoring International
  Warehouse Operations in Europe, Production and
  Operations Management, 17(2), 175-183.
• McGinnis, L.F., A. Johnson, and M. Villarreal
  (2006) Benchmarking Warehouse Performance Study,
  Technical Report, Georgia Institute of
  Technology.

29
de Koster and Balk (2008)

• inputs
• of direct FTEs
• size of the W/H
• degree of automation
• of SKUs

• outputs
• of order lines picked/day
• level of value-added logistics (VAL) activities
• of special optimized processes
• of error-free orders shipped out
• order flexibility

30
de Koster and Balk (2008)

• 65 warehouses containing 140 EDCs
• EDC distribution centers in Europe responsible
  for the distribution for at least five countries
  there
• composition
• results

31
Warehouse Performance Study in GIT

• develop a single index to measure the performance
  of a warehouse
• use data envelope analysis

32
Examples from the Index Warehouse Size

• What are your inferences?

33
Examples from the Index Mechanization

• What are your inferences?

34
Profiling?? Examples Only
35
Profiling

• profile of the warehouse
• define processes
• status of processes
• reveal status of warehouse
• purposes
• get new ideas on design and planning
• get improvement
• get baseline for any justification
• remarks
• use distributions, not means
• express in pictures

36
Various Profiles

• indicators on every aspect
• receiving, prepackaging, putaway, storage, order
  picking, packaging, sorting, accumulation,
  unitizing, and shipping

37
Customer Order Profiling
Customer Order Profile
results from order profiling help design a
warehouse, including its layout, equipment,
picking methods, etc.
38
Family Mix Distribution

• implication zoning by family

39
Handling Unit Mix Distribution Full/Partial
Pallets

• implication good to have a separate picking area
  for loose cartons

40
Handling Unit Mix Distribution Full/Broken
Cases

• implication good to have a separate picking area
  for broken cases

41
Order Increment Distributions - Pallets

• implication good to have ¼ and ½ pallets

42
Order Increment Distributions - Cases

• implication good to have ½-size cases

43
Lines per order Distribution

• implication on the picking methods

44
Lines and Cube per order Distribution

• implication on the picking methods

45
Items Popularity Distribution

• implication on storage zones, golden, silver,
  bronze

46
Cube-Movement Distribution

• implication small items in drawers or bin
  shelling large items in block stacking,
  push-back rack

47
Popularity-Cube-Movement Distribution

• implication on storage mode

48
Item-Order Completion Distribution

• implication on mode of storage, e.g., warehouse
  within a warehouse

49
Demand Correlation Distribution

• implication on zoning of goods

50
Demand Variability Distribution

• implication variance of demand to set safety
  stock

51
Item-Family Inventory Distribution

• implication area assigned to different types of
  storage

52
Handling Unit Inventory Distribution

• implication different storage modes according to
  the number of pallets on hand

53
Seasonality Distribution

• implication shifting human resources and
  possibly space

54
Daily Activity Distribution

• implication shifting human resources and
  possibly space

55
Activity Relationship

• implication on layout