1. IE7201: Production & Service Systems Engineering Spring 2017
Instructor: Spyros Reveliotis
homepage:

2. “Course Logistics” Office Hours: By appointment Course Prerequisites:
ISYE 6761 (Familiarity with basic probability concepts and Discrete Time Markov Chain theory)
ISYE 6669 (Familiarity with optimization concepts and formulations, and basic Linear Programming theory)
Grading policy:
Homework: 0%
Two Midterm Exams: 30% each
Final Exam: 40%
Reading Materials:
Course Textbook: Fundamentals of Queueing Theory, by D. Gross, J. G. Shortle, J. M. Thompson and C. M. Harris, J. Wiley & Sons, Inc., 2008.
Additional material will be distributed during the course development

3. Course Objectives
Provide an understanding and appreciation of the different resource allocation and coordination problems that underlie the operation of production and service systems.
Enhance the student ability to formally characterize and study these problems by referring them to pertinent analytical abstractions and modeling frameworks.
Develop an appreciation of the inherent complexity of these problems and the resulting need of simplifying approximations.
Systematize the notion and role of simulation in the considered problem contexts.
Define a “research frontier” in the addressed areas.

4. Our basic view of the considered systems
Production System: A transformation process (physical, locational, physiological, intellectual, etc.)
Organization
Inputs
Outputs
Materials
Capital
Labor
Manag. Res.
Goods
Services
The production system as a process network
Stage 5
Stage 4
Stage 3
Stage 2
Stage 1
Suppliers
Customers

5. The major functional units of a modern organization
Strategic Planning:
defining the organization’s mission and
the required/perceived core competencies
Production/
Operations:
product/service
creation
Finance/
Accounting:
monitoring of
the organization
cash-flows
Marketing:
demand
generation
and
order taking

6. Fit Between Corporate and Functional Strategies (Chopra & Meindl)
Corporate Competitive Strategy
Product
Development
Strategy
Supply Chain
or Operations
Strategy
Marketing
and Sales
Strategy
Information Technology Strategy
Finance Strategy
Human Resources Strategy

7. Corporate Mission The mission of the organization
defines its purpose, i.e., what it contributes to society
states the rationale for its existence
provides boundaries and focus
defines the concept(s) around which the company can rally
Functional areas and business processes define their missions such that they support the overall corporate mission in a cooperative and synergistic manner.

8. Corporate Mission Examples
Merck: The mission of Merck is to provide society with superior products and services-innovations and solutions that improve the quality of life and satisfy customer needs-to provide employees with meaningful work and advancement opportunities and investors with a superior rate of return.
FedEx: FedEx is committed to our People-Service-Profit philosophy. We will produce outstanding financial returns by providing totally reliable, competitively superior, global air-ground transportation of high-priority goods and documents that require rapid, time-certain delivery. Equally important, positive control of each package will be maintained utilizing real time electronic tracking and tracing systems. A complete record of each shipment and delivery will be presented with our request for payment. We will be helpful, courteous, and professional for each other, and the public. We will strive to have a completely satisfied customer at the end of each transaction.

9. A strategic perspective on the operation of the considered systems
Responsiveness (Reliability; Quickness; Flexibility;
e.g., Dell, Overnight Delivery Services)
Competitive Advantage through which
the company market share is attracted
Cost Leadership (Price;
e.g., Wal-Mart, Southwest
Airlines, Generic Drugs)
Differentiation (Quality; Uniqueness;
e.g., Luxury cars, Fashion Industry,
Brand Name Drugs)

10. The operations frontier, trade-offs, and the operational effectiveness
Responsiveness
Cost Leadership
Differentiation

11. The primary “drivers” for achieving strategic fit in Operations Strategy (adapted from Chopra & Meindl)
Corporate Strategy
Operations Strategy
Efficiency
Responsiveness
Market
Segmentation
Facilities
Inventory
Transportation
Information

12. The course perspective: Modeling, analyzing and controlling workflows
Some Key Performance measures
Production rate or throughput, i.e., the number of jobs produced per unit time
Production capacity, i.e., the maximum sustainable production rate
Expected cycle time, i.e., the average time that is spend by any job into the system (this quantity includes both, processing and waiting time).
Average Work-In-Process (WIP) accumulated at different stations
Expected utilization of the station servers.
Remark: The above performance measures provide a link between the directly quantifiable and manageable aspects and attributes of the system and the primary strategic concerns of the company, especially those of responsiveness and cost efficiency.

13. The underlying variability
But the actual operation of the system is characterized by high variability due to a large host of operational detractors; e.g.,
machine failures
employee absenteeism
lack of parts or consumables
defects and rework
planned and unplanned maintenance
set-up times and batch-based operations

14. Analyzing a single workstation with deterministic inter-arrival and processing times
Case I: ta = tp = 1.0 TH B1 M1
WIP 1
TH = 1 part / time unit
Expected CT = tp t 1 2 3 4 5
Arrival
Departure

15. Analyzing a single workstation with deterministic inter-arrival and processing times
Case II: tp = 1.0; ta = 1.5 > tp TH B1 M1
WIP
Starvation! 1
TH = 2/3 part / time unit
Expected CT = tp t 1 2 3 4 5
Arrival
Departure

16. Analyzing a single workstation with deterministic inter-arrival and processing times
Case III: tp = 1.0; ta = 0.5 TH B1 M1
WIP t 1 2 3 4 5
Arrival
Departure
Congestion!
TH = 1 part / time unit
Expected CT  

17. A single workstation with variable inter-arrival times
Case I: tp=1; taN(1,0.12) (ca=a / ta = 0.1) TH B1 M1
WIP 3 2
TH < 1 part / time unit
Expected CT   1 t 1 2 3 4 5
Arrival
Departure

18. A single workstation with variable inter-arrival times
Case II: tp=1; taN(1,1.02) (ca=a / ta = 1.0) TH B1 M1 t 1 2 3 4 5
Arrival
Departure
WIP
TH < 1 part / time unit
Expected CT  

19. A single workstation with variable processing times
Case I: ta=1; tpN(1,1.02) TH B1 M1 t 1 2 3 4 5
WIP
TH < 1 part / time unit
Expected CT  
Arrival
Departure

20. Remarks
Synchronization of job arrivals and completions maximizes throughput and minimizes experienced cycle times.
Variability in job inter-arrival or processing times causes starvation and congestion, which respectively reduce the station throughput and increase the job cycle times.
In general, the higher the variability in the inter-arrival and/or processing times, the more intense its disruptive effects on the performance of the station.
The coefficient of variation (CV) defines a natural measure of the variability in a certain random variable.

21. The propagation of variabilityW1 W2 B1 M1 TH B2 M2
Case I: tp=1; taN(1,1.02)
Case II: ta=1; tpN(1,1.02)
WIP t 1 2 3 4 5
WIP 3 2 1 t 1 2 3 4 5
W1 arrivals
W1 departures
W2 arrivals

22. Remarks
The variability experienced at a certain station propagates to the downstream part of the line due to the fact that the arrivals at a downstream station are determined by the departures of its neighboring upstream station.
The intensity of the propagated variability is modulated by the utilization of the station under consideration.
In general, a highly utilized station propagates the variability experienced in the job processing times, but attenuates the variability experienced in the job inter-arrival times.
A station with very low utilization has the opposite effects.

23. Some key issues to be addressed in this course
How do I get good / accurate estimates of the performance of a certain system configuration?
How do I design and control a system to support certain target performance?
What are the attributes that determine these performance measures?
What are the corresponding dependencies?
Are there inter-dependencies between these performance measures and of what type?
What target performances are feasible?

24. Queueing Theory: A plausible modeling framework
Quoting from Wikipedia:
Queueing theory (also commonly spelled queuing theory) is the mathematical study of waiting lines (or queues).
The theory enables mathematical analysis of several related processes, including arriving at the (back of the) queue, waiting in the queue (essentially a storage process), and being served by the server(s) at the front of the queue.
The theory permits the derivation and calculation of several performance measures including the average waiting time in the queue or the system, the expected number waiting or receiving service and the probability of encountering the system in certain states, such as empty, full, having an available server or having to wait a certain time to be served.

25. Factory Physics (a term coined by W. Hopp & M. Spearman)
The employment of fundamental concepts and techniques coming from the area of queueing theory in order to characterize, analyze and understand the dynamics of (most) contemporary production systems.

26. Automation and the need for behavioral control
J1 : R1 ® R2 ® R3
J2 : R3 ® R2 ® R1
Explain the system operation and the depicted deadlock
Emphasize that deadlock is a disruption, in general, and a “fatal error” in case of automated systems

27. Cluster Tools: An FMS-type of environment in contemporary semiconductor manufacturing
Explain the basic structure and operation
Cassettes of wafers enter the cluster tool at a load port, where they are opened
Wafers are processed one at a time in chambers mounted to the tool
Part transfer is supported by a centrally located robotic manipulator
Current approach to deadlock resolution very conservative:
All chambers of the same type (i.e., parallel processing)
All wafers follow the same acyclic sequence of stages (i.e., flow line)
Net result: lack of flexibility

28. Another example: Traffic Management in an AGV System
Explain zone control for vehicle collision
Discuss the types of occurring deadlocks, emphasizing Type-2 deadlock

29. A more “realistic” example: A typical fab layout
Prevailing practice for avoiding Type-2 AGV deadlock in contemporary fabs: establishment of unidirectional traffic loops (TANDEM AGV’s)
Potential problems:
Unnecessarily long routes
A slow / stopped vehicle can block the entire traffic

30. A more avant-garde example: Computerized workflow management
The basic idea: Deploy computer-based (frequently, Web-enabled) software that monitors, coordinates and controls the execution of various business transactions that take place in a sequence of stages (such an approach is deemed especially appropriate for repetitive / routine processes taking place in banking, insurance claim processing, e-commerce, etc.)
The resulting software is known as Workflow Management Systems
Major components of WfMS:
An interface for specifying the process sequencing logic and the involved resource allocation patterns
A controller that will monitor and enforce the progress of the various enacted cases
Extensions of this idea can be applied in internet-based computing / supporting process execution in computational grids

31. An example taken from the area of public transportation

32. A modeling abstraction: Sequential Resource Allocation Systems
A set of (re-usable) resource types R = {Ri, i = 1,...,m}.
Finite capacity Ci for each resource type Ri.
a set of job types J = {Jj, j = 1,...,n}.
An (partially) ordered set of job stages for each job type, {pjk, k = 1,...,lj}.
A resource requirements vector for each job stage p, ap[i], i = 1,...,m.
A distribution characterizing the processing time requirement of each processing stage.
Protocols characterizing the job behavior (e.g., typically jobs will release their currently held resources only upon allocation of the resources requested for their next stage)
Sequential RAS: A unifying modeling framework for addressing the behavioral problems encountered in all the previous cases

33. Behavioral or Logical vs Performance Control of Sequential RAS
Resource
Allocation
System
Behavioral
Correctness
Efficiency
Traditionally, resource allocation systems have been studied in terms of their efficiency
Here we are also concerned with their behavioral correctness

34. An Event-Driven RAS Control Scheme
System State Model
Logical Control
Performance Control
Event
Feasible
Actions
Admissible
Actions
Commanded
Action
A closed-loop control scheme:
Controller actions are responses to the events taking place in the RAS
Situation assessment based on a maintained RAS state model gives the feasible actions
A logical control policy filters out the admissible actions
A performance control policy selects the admissible action to be commanded on the system
Configuration Data
RAS Domain

35. Theoretical foundations
Control
Theory
“Theoretical”
Computer
Science
Discrete
Event
Systems
Control theory: since we are dealing with a control problem and offers the fundamental perspectives of state, state-based control, feedback, etc.
Theoretical computer science: the main paradigm for modeling and analyzing behaviors
Operations Research: for modeling and addressing performance, but also for designing logical control policies
Operations
Research

36. Course Outline
1. Introduction: Course Objectives, Context, and Outline
Contemporary organizations and the role of Operations Management (OM)
Corporate strategy and its connection to operations
The organization as a resource allocation system (RAS)
The underlying RAS management problems and the need for understanding the impact of the underlying stochasticity
The basic course structure
Modeling and Analysis of Production and Service Systems as Continuous-Time Markov Chains
A brief overview of the key results of the theory of Discrete-Time Markov Chains
Bucket Brigades
The Exponential Distribution and the Poisson Process
Continuous-Time Markov Chains (CT-MC)
Birth-Death Processes and the M/M/1 Queue
Transient Analysis
Steady State Analysis
Modeling more complex behavior through CT-MCs
Single station systems with multi-stage processing, finite resources and/or blocking effects
Open (Jackson) and Closed (Gordon-Newell) Queueing networks

37. Course Outline (cont.) 3. Accommodating non-Markovian behavior
Phase-type distributions and their role as approximating distributions
The M/G/1 queue
Priority Queues
The G/G/1 queue
The essence of “Factory Physics”
(Reversibility and BCMP networks)
4. Performance Control of Production and Service systems
Controlling the “event rates” of the underlying CT-MC model (an informal introduction of the dual Linear Programming formulation in standard MDP theory)
A brief introduction of the theory of Markov Decision Processes (MDPs) and of Dynamic Programming (DP)
An introduction to Approximate DP
An introduction to dispatching rules and classical scheduling theory
Buffer-based priority scheduling policies, Meyn and Kumar’s performance bounds and stability theory

38. Course Outline (cont.)
5. Behavioral Control of Production and Service Systems
Behavioral modeling and analysis of Production and Service Systems
Resource allocation deadlock and the need for liveness-enforcing supervision (LES)
Petri nets as a modeling and analysis tool
A brief introduction to the behavioral control of Production and Service Systems