Advanced Planning and Scheduling

1
Advanced Planning and Scheduling

• References
• Errington, APS - a powerful emerging
  technology, 1997
• i2 technology product information
• Adexa product information
• i2 FP training materials

2
Outline

• Introduction to APS
• Implementation of APS
• APS solutions
• Example of APS planning engine - FP

3
Traditional Divided Process
Prod./Res. Planning
MASTER
Demand
M. Sched. RCCP
MPS
MRP CRP
DRP
EXECUTION
DISTRIBUTION
Detailed Scheduling
Demand Mangmt.
4
Divisions of the Complex Problem

• Division by planning time horizon
• long-term/strategic, medium-term/tactical, and
  short-term/operational
• Division by different constraint concerns
• material constraint / capacity constraint
• manufacturing constraint / distribution
  constraint
• Division by different detail levels
• master production scheduling / detailed
  scheduling
• strategic capacity planning / detailed resource
  planning
• demand promising / due date quote

5
Deficiencies of the Traditional Process

• Divisions to simplify the software
• complexity is minimized by breaking up the
  problem
• local optimums do not sum up to a global optimum
• burden is shifted to human planner
• Divisions lead to iteration
• only part of the whole problem is considered by
  each planning tool
• plan results and problems are propagated among
  tools
• iteration is needed to achieve a better plan
  quality
• Manual iteration is impractical
• manual effect analyses of changes made
• manual judgement to make changes

6
Drivers of New Planning System

• Computing power has changed
• hardware capability has advanced dramatically
• software design has advanced significantly
• friendly user graphical interfaces
• Planning problem has changed
• 1980s competitive drivers - TQM and JIT
• 1990s competitive drivers - responsiveness and
  flexibility
• accurate, real-time, dynamic planning
  capabilities required
• Planning process must change
• BPR/TOC
• enhancing the competitiveness instead of
  automating the business process

7
Supply Chain Optimization / Foundation of
eBAdvanced Planning and Scheduling (APS)

• Is a redesign of the overall planning process
  that integrates the process across several
  dimensions
• Addresses the whole planning problem directly,
  rather than dealing with small parts of the
  problem separately

8
Concurrent Planning
Demand Forecast / Customer Orders
Available Production Capacity
Raw Materials Procurement / Supplier Lead Times
Planning Engine
Resources
Customer Due Dates
Inventory Levels
Distribution/ Warehouse Routing
Optimal Plan
9
Integrated Planning
10
The Integration of APS

• Integrating master and execution planning
• Integrating material and capacity
• Integrating manual and automated planning
• Making order promising dynamic and real-time

11
What is New in the APS

• Memory-based processing and advanced algorithms
  to compute sophisticated production and
  distribution plans that consider multiple
  constraints
• Synchronize constrained resources and materials
  flow to a detailed level
• Quick re-plans by including the latest
  information
• Rule-based logic allows users to capture their
  business and production strategies and priority
• What-if analysis to support decision making
  process

12
Focuses of APS

• Supply chain planning
• Include some features of MRPII ERP
• Develop plan for multiple facilities, suppliers,
  and logistics
• Goal determine what to make and where to make,
  not individual operations
• Multi-plant scheduling
• Allocate order over multiple facility and develop
  schedule
• Plant scheduling
• Develop detailed operation or job schedules
• Consider both materials and components
  requirements

13
Approaches of APS

• Constraint-based planning
• Heuristic engine rely on the forward and
  backward scheduling to balance orders against the
  resources
• Theory of constraints engine identify resource
  bottlenecks and then prioritize activities to
  minimize the bottlenecks
• Simulation engine model shop floor, process
  local heuristic scheduling rule, and generate
  statistical output about the flow through work
  centers
• Knowledge-based engine use implicit rules about
  customer demand, work flow, resources, and
  constraints to balance incoming orders against
  delivery dates

14
Approaches of APS (contd)

• Optimizer engine use mathematical programming
  and branching techniques to minimize scheduling
  and resource conflicts in meeting individual
  customer orders
• Network-based planning
• Work from top-down starting from customer order
• Schedule one customer order at a time, resolve
  materials and capacity issues concurrently on
  every operations

15
Outline

• Introduction to APS
• Implementation of APS
• APS solutions
• Example of APS planning engine - FP

16
Selecting the Software

• Criteria
• Focus distribution? manufacturing? type of
  industries?
• Maturity and support
• Feature all key system constraints considered?
• Integration easy? Seamless?
• Technology latest software technology used?
• Platform flexible?
• Price of revenue?
• Best way use a small, but representative, set of
  data including all the key elements, then ask
  vendors to create prototypes for demo

17
Implementation of APS

• Business reengineering (?)
• Develop a model of manufacturing system using the
  constructs provided by the packages
• Integrate with the existing systems and data
  cleaning
• Customize
• Including second-party reporting software,
  special rules, nonstandard feature
• Train users and individuals for maintaining the
  system

18
Interactions between ERP/APS and MES
19
Issues of Implementing APS Solutions

• APS ready? sound enterprise information system
  required (not necessary an enormous ERP system)
• APS solution is not a software package, its a
  way of doing business and its a universal way
  (everyone including your competitors).
• Are your competitive advantage compromised by
  implementation of SCP solutions?

20
Outline

• Introduction to APS
• Implementation of APS
• APS solutions
• Example of APS planning engine - FP

21
Gartner Process Industry Gartner Group Magic
Quadrant in Oct.99
Challengers
Leaders
Manugistics
Ability to Execute
SCT/Fygr
i2
JD Edward
Logility
Adexa
SAP
LPA
Mercia
Aspen Tech
Demand Mgmt
As of 9/30/99
Niche Players
Visionaries
Completeness of Vision
22
Gartner Discrete Industry Gartner Group Magic
Quadrant in Oct.99
Challengers
Leaders
i2
Ability to Execute
Manugistics
Synquest
Baan
Adexa
Peoplesoft
STG
Logility
Web Plan
LPA
Thru-Put
Mercia
Edwards
JD
Aspen Tech
Demand Mgmt
As of 9/30/99
Niche Players
Visionaries
Completeness of Vision
23
i2 TradeMatrix Solutions Overview
DesignPartners
Direct Material Suppliers
ProductDevelopment
Product Development/Engineering
Customers
Customer Management
Supply Chain Planning
Procurement
Manufacturing
Service Management
Marketing / Sales / Administration
Indirect Material Suppliers
Fulfillment
Logistics Providers
24
i2 TradeMatrix Solutions

• For product development Design Solution
• collaborated design and development for low cost,
  low risk and short time-to-market
• For procurement Buy Solution
• efficient sourcing, negotiation, collaboration
  and ordering
• For supply chain planning Plan Solution
• Efficient conversion of raw material into
  customer-ready product offerings
• For customer management Sell Solution
• marketing, selling, customer collaboration, order
  processing and monitoring
• For order promising and logistics Fulfill
  Solution
• intelligently commitments to customer order and
  requests
• For service planning and scheduling Service
  Solution
• Increase of customer satisfaction and revenue
  with minimum investment

25
i2 TradeMatrix Plan Solutions
tactical
strategic
operational
execution
strategic businessnetwork planning
messaging and data integration
buy
manufacturing planning
command control
scheduling
make
move
transportation
configu ration
store
distribution/inventory planning
sell
demand planning/ inventorysop planning
26
i2 TradeMatrix Plan Solution
Portfolio Planning
Long Term Forecast
Transition Triggers
Strategic Planning - Capacity Expansion - Network
Planning - Inventory Planning - Capacity
Allocation
Capacity Plan
Constrained Forecast
Optimal Portfolio
Transition Planning
Transition Plans
Capacity Allocations
Master Planning - Demand/Supply Match - Logistics
Optimization - Allocations Management
Customer Management - eCommerce - eConfig -
eMarketing - eCare - Collaboration
Demand Fulfillment - Order Promising - Orders
Fulfillment - Internet Fulfillment - Allocations
Mgmt. - Forecast Netting
Demand Plan
Allocations ATP
Demand Planning
Collaboration
Netted Forecast
Orders (New/Changed)
Collaboration
Actual Starts
Order Status
Request Starts/Outs
Factory Planning
Factory Planning
MES Data
Lot Due Dates
Real Time Dispatching
MES
Order Entry
27
Adexa iCollaborationTM Solution
Business Consumer
Channel Intermediaries
CDP
CSP
GSP
Direct Channel
Home/Business Consumer
SCP
GRA
Channel Resellers
Home Consumer
Aggregators Metamediary
28
Adexa iCollaboration Suite
Strategic
PDP - Product Development Planner
GSP - Global Strategic Planner
SCP - Supply Chain Planner
CSP - Collaborative Supply Planner
CDP - Collaborative Demand Planner
MCP - Material and Capacity Planner
Single Data Model
Business Intelligence
RDS - Reactive Dynamic Scheduler
GRA - Global Real-Time ATP
CEP - Collaborative Enterprise Planner
Operational
SCC - Supply Chain Controller
Supply Chain Execution
Enterprise Resource Planning
Product Design Management
Customer Relationship Management
Product Configurator
Transaction Systems
29
Adexa Supply Chain Solution
30
Solution Components of i2 and Adexa
31
Strategic Planning

• Profit optimization - maximizing revenue while
  minimizing cost
• Capacity expansion and allocation
• efficient capacity investment
• capacity balancing
• most profitable product mix
• Distribution network planning
• determination of subcontracting, sourcing, and
  other alternatives
• locations and capacity for distribution centers,
  plants, and other facilities based on projected
  demand
• What-if scenario simulation
• stochastic (Monte Carlo) simulation by i2

32
Issues of Demand Planning

• Most unreliable info in planning activities
  demand forecast
• Bullwhip effect fluctuation propagates and
  magnified through supply chain
• Risk pooling aggregating demand for less
  fluctuation of demand forecast but how?
• Different perspectives of demand customer,
  technology, sales region, etc
• Uses of demand forecast for material planning?
  or capacity allocation?

33
Demand Planning

• Statistical forecast techniques are standard
• Multi-dimensional slice-and-die analysis to
  provide different views for various divisions of
  the company on-line analytic processing (OLAP)
  technology

34
Demand Promising

• ATP (available-to-promise) and
• CTP (capable-to-promise)
• Access to delivery dates and order/quote status
• Reservation of inventory and planned production
  for new orders
• Distributed, global, real time
• Can match supply and demand based on options
• Can trigger re-plan at SCP/MCP level

35
ATP and AATP

• ATP
• the uncommitted portion of inventory
• the end item supply that can be used to
  quote/promise/reschedule orders
• Traditional ATP practice (FCFS)
• low margin demand consume all available resources
• nothing available for emergency order or high
  priority customer
• AATP Allocated Available to Promise
• AATP is ATP allocated by different demand
  perspectives
• AATP provides a mechanism to reserve supply
• Allocation rules need to be defined by users

36
Capable to Promise (CTP )

• Extend the end item representation to include
  multiple levels of BOM as modeled in MP
• Provide a real time mechanism to promise order by
  searching
• End item
• Raw material / WIP considering capacity
  constraints
• alternate routings
• alternate parts
• alternate resources

37
Enterprise Integrated Planning

• Enterprise-wide planning
• Simultaneous material, manufacturing, and
  distribution planning
• Synchronizes planning for multiple plants and
  other facilities, dynamically managing
  inter-plant dependencies
• User-defined attributes and business rules
• Integration with lower levels of planning.
• Enterprise-wide what-if analysis
• Synchronization with demand planning
• Synchronization with demand promising

38
Manufacturing Planning

• Plans for manufacturing operations subject to
  both material and capacity constraints
• MCP(Materials and capacity planning) vs.
• FP(Factory planner)
• MCP balancing and scheduling
• FP infinite capacity planning(ICP) and finite
  capacity planning(FCP)
• RDS (Reactive dynamic scheduling) vs. FP
• RDScontinuous capacity profile, sequence
  dependent setups, batch resources, secondary
  resources
• FP detailed scheduling

39
Outline

• Introduction to APS
• Implementation of APS
• APS solutions
• Example of APS planning engine - FP

40
FP - Concept
Step -1 Run Infinite Capacity Plan a. Will
identify Material and Capacity
problems b. Fix Material problems Step - 2 Run
Finite Capacity Plan - CAO a. Will attempt to
fix Capacity problems Step - 3 Run
Advanced Scheduler
41
FP - Planning Flow
Construct Manufacturing Model
to identify critical constrained resources
ICP
FCP (CAO)
to synchronize the flow
Adv. Schedule/Finish
42
Infinite Capacity Planning
Inventory Assignment
Backward Schedule
Forward Schedule
Planned Start Time
43
Inventory Assignment

• Step 1 Sort Demand Orders
• Due Date
• Priority Points (the more, the earlier)
• Quantity (the smaller, the earlier)
• These factors can be considered in any order

44
Inventory Assignment - Example
45
Inventory Assignment

• Step 2 BOM (bill of materials) explosion

PART SEQUENCE
A
B
C
D
46
Inventory Assignment

• Step 3 Assign Inventory
• Iterate through part sequence and post demands
• Assign inventory to each demand order
• Create Manufacturing Orders
• Planned Quantity Required Quantity - Total
  Inventory - WIP

47
Inventory Assignment - Example
48
Inventory Assignment - Example (contd)
49
Backward Schedule

• Definition Latest Possible Start Time (LPST)
• the latest date a manufacturing task can begin
  and still complete on time
• Calculation
• LPST Due Date - (SetupRunCool Down)
• Each manufacturing task will have an LPST

50
Example
LPST
Table Top
Order_1-MFG00002
Table
Table Assembly
LPST
Order_1-MFG00000
Order_1 Due Date 3/15/95
LPST
Legs
Order_1-MFG00001
Backward Scheduling
51
Forward Schedule

• Definition Earliest Possible Start Time (EPST)
• The earliest date that a task may begin
• Calculation
• EPST max (Server Start date , Material
  availability date)
• Each task will have an EPST

52
Example
EPST
Table Top
CompletionDate
EPST
Table Assembly
EPST
Legs
Time Propagation
53
Planned Start Time

• Definition Planned Start Time (PST)
• is the expected start date that a task will begin
  taking into consideration material availability
• PST max (EPST , LPST)
• PST values may be overridden through the user
  interface

54
Example
LPST
Table Top
PST, EPST
Table Top
LPST
Due Date
Table Assy.
Table Assy.
LPST
Legs
CompletionDate
PST, EPST
PST, EPST
Legs
WIP
55
Finite Capacity Planning

• Manual Balancing
• Increase capacity on overutilized resource if
  possible
• Interactively move orders
• Interactively done orders
• Automated Balancing
• Constraint Anchor Optimization (CAO)

56
CAO - Concept
Step - 1 Identify Overloaded Resources -
Capacity Shortages Step - 2 Pick an overloaded
resource and balance by pulling - so the order
is not LATE offloading - so the order is not
LATE pushing - has no other way Step - 3 If
balanced select the next overloaded resource
and repeat above steps
57
Capacity Problem Identification
Step - 1 Identify Overloaded Resources -
Capacity Shortages 1. What if there are more
than ONE resource that is overloaded ? 2.
How is CAO going to prioritize ?
Resource Criticality A value assigned to each
overloaded resource that reflects the average
number of tasks affected by this resource.
58
Resource Criticality
M1
M2
M2 will have higher Criticality value than M1
59
Tasks in the Critical Resource
Step - 2 Once resource identified it will start
with the earliest overloaded bucket Which
tasks among in the bucket to consider for
moving ?
60
Task LPST
CAO Priority All tasks in a bucket are assigned
a value based on LPST.
1
3
2
LPST
61
Task Priority
CAO Priority LPST - start_time
62
Pull Tasks - First Pass
Pull-Offload-Push
2
1
3
1. Compute total available capacity in earlier
buckets 2. Pull the task that has a total runtime
less than the total available capacity
based on CAO Priority
63
Next Overloaded Bucket
Pull-Offload-Push
1
3
3. The resource overload in bucket 4 solved.
So go to next overloaded bucket ie., 5
64
Continue Pulling
Pull-Offload-Push
1
3
65
Finish of First Pass
Pull-Offload-Push
1
3
4. End of bucket reached. Repeat procedure
from start of the bucket
66
Pull - Second Pass
Pull-Offload-Push
1
67
Second Pass (contd)
Pull-Offload-Push
68
Offload and Push - Third Pass
Pull-Offload-Push

• Cannot pull any tasks due to MATERIAL CONSTRAINT
• OFFLOAD if alternate exists
• Push - Higher CAO Priority pushed first

69
Lateness Minimization
Why higher CAO Priority during Push ? gt Minimize
LATENESS
1
3
2
LPST
70
Imposing Limits on Planning

• Consider the following.
• 1. How many times to scan the buckets?
• cao_reset_limit 3000 (server flag)
• 2. How many times a task can be pulled-pushed?
• Change in direction
• convergence_speed 4 (specified using UI)

71
CAO - Resource Criticality
STEP - 3 Once cao_reset_limit reached it
propagates the changes and recomputes the
resource criticality values. Perform
Pull-Offload-Push (Repeat the procedure) for the
next overloaded resource
72
Resource Cycles
Issues to consider.Same Sequence of resource are
balanced A CYCLE is created
Balance Limit Limits the number of times a given
resource is balanced