1 RESEARCH AND TEACHING PLAN Presentation to: Department of Information and Decision Sciences College of Business Administration University of Texas at.

1

RESEARCH AND TEACHING PLAN

Presentation to:

Department of Information and Decision Sciences

College of Business Administration

University of Texas at El Paso

2

PRESENTATION OUTLINE

• Describe an “interesting” research project

• Present an overview of current research and teaching

interests/projects

• Discuss my vision for the future

3

“Interesting” Research Project:

KABOMKABOMknowledge/activity based operations knowledge/activity based operations

modelingmodeling

WithCarey McCleskey, Edgar Zapata, Russ

Rhodes

Advanced Projects, NASAKennedy Space Center, FL

4

Outline for this section

• Introduction• Previous work

• KABOM• Project Extensions

• Project Status

5

Introductiondo you know?

• The cost per space shuttle flight?

• During the design of shuttle, the designers expected/planned the time between flights to be:

a) < 1 weekb) < 1 monthc) 1 - 2 monthsd) 3 - 4 monthse) > 5 months

• From the above, what is the current average time between flights?

• What is the shuttle’s design reliability?

6

Introductionthe design to operations gap

• A view of operations by the designers

• Operations at the Orbiter Processing Facility (OPF) today.

7

Introductionthe causes

• Space vehicles are complex systems• Operations was not a primary concern of

designers– Design process focused on the rocket equation– Primary operations concern: how do we use existing

facilities– Once the vehicle worked - up to Kennedy to make it

work.

• Minimizing cost was an objective (NASA’s budget has had its ups and downs), BUT it was a “myopic” view, minimize development cost or minimize manufacturing cost, seldom a total cost view.

8

Introductionthings are changing• NASA is changing.

– Moving from a service provider to a support and monitoring entity

• Space business environment is changing– Commercial launch capabilities have increased.

• Facilities in old USSR, Guyana, Brazil, China, and even sea launches

– Need for launches has increased, exceeding capacity

– Interest in other commercial uses of space• Hilton Space Hotels?- spacehotels

– The X Prize

9

Outline for this section

• Introduction• Previous work

• KABOM• Project Extensions

• Project Status

10

Previous workvision spaceport

• Team approach > ground operations modeling– Includes NASA KSC and Ames, Lockheed Martin,

Boeing, SIAC, and other related government, university, and private organizations.

– My role: NASA fellow for the summers of 1998 and 1999.

• Objectives:– Develop models to asses new designs: spaceport

operations - life cycle costs– Educate vehicle designers on spaceport operations– Connect model to design/development models and

manufacturing models

11

Previous workvision spaceport• First approach

– Knowledge based utility functions– Mapped vehicle inputs to spaceport

functions. • Functions provided sense of direction: was the design

an improvement over the baseline for a function. Cost and times derived from the deviation

• Developed a tool in Excel/VB: AATe. – NASA Copyright. – Used in several studies at Marshal and

Langley.– Modeling approach was applied in a second

tool - application developed by CCT.

12

F(Mc)=i1.I, j1..Ji (Xij*SiMc/(y =1..ISyMc))

I # of design/assessment questions Ji # of options for design question iXij Value of option j for design question i (0 - 1)SiM Strength of relation, design/assessment question i to McF(Mc) Score for spaceport function M and characteristic c

$ or CT

10

Umin

Umax

Previous workvision spaceport

13

Outline for this section

• Introduction• Previous work

• KABOM• Project Extensions

• Project Status

14

Knowledge/Activity based Operations Modeling

(KABOM)• Objective:

– To asses the operational requirements of a vehicle design. Determine:

• Cost per flight (cost per pound) and Cycle time• Processes and Resources required

– Present to vehicle designers the processes required for operation

– Address limitations of the utility function approach• Dissipate effect an expert will have captures• Difficult to calibrate as inputs increase• Limited representation of reliability effect

15

KABOM BasicsOrigins

• Related to ABC models - used in manufacturing and SCM

• Similar to the approach used by Christenson and Komar (1998) to model/ analyze reusable rocket engine operability

• Primarily for online functions (landing, turnaround,..)

• Model is a component of total spaceport analysis model which should capture the other spaceport modules (not in flow).– This is a subset of an LCC model that includes

development, manufacturing, and operations

16

KABOM BasicsThe RLV design

• The model characterizes a RLV architecture/concept by I design variables– I represents a particular option of the

vehicle, for example ceramic times, ultra high density ceramics, SOFI.

– The binary variable di is used to represent the inclusion of a design option; di = 1 if the design option is included in the design and di = 0 if not.

17

KABOM BasicsThe RLV design

• This model characterizes a RLV architecture/concept by J vehicle characteristics/ operational drivers. – The vehicle characteristics J represents

measures that will drive operational cost or time, for example the number of fuel cells or the area covered by a type of thermal protection.

– The variable qj is used to represent the quantify of an operational driver, qj > 0 if that operational driver exists in the design and qj = 0 if not.

18

KABOM BasicsThe Activities

• There are A spaceport activities.– Activities relate to one or more design variables.

– The binary variable sa represents the inclusion of that activity in the activity set for that design; sa = 1 if that activity is part of the activity set and sa = 0 if it is not.

– The determination of the activity set is determined by knowledge based equations. I.E.:s3 = 1 , if d1 + d12 = 2

0, otherwises7 = 1 , if d11 = 0 and d123 = 1 or if q56 >

1000, otherwise

19

KABOM BasicsActivity Characterization

• Three characteristics for each activity: activity time, activity cost, and need percentage.– The process time for an activity a; pa, is

determined by a knowledge based equation. • p1 = UNIF (35, 100) x q13 hours

• p5 = EXPO (3 x q38 ) minutes

• p42 = 50 x q38 minutes

– The cost per activity a; ca, will be characterized in a similar fashion.

• c1 = p1 x $14,000

• c81 = $10,000 + $120 x q39

• c22 = $75,000

20

KABOM BasicsActivity Characterization

• Continued:– The need percentage for an activity a; na, is

determined by a knowledge based equation. • n19 = 10% , if e45 = 1

30% , if e46 = 1

100%, otherwise

• n7 = 50% , if q92 < 2,000

100%, otherwise

– From these values, we can determine the expected average time for an activity, E’(pa):

E’(pa) = na x E(pa)

21

KABOM BasicsProcess Modeling

• Spaceport process model– All A activities are integrated into a

spaceport process model. – The process model captures precedence

constraints and possible alternative routes.

LAUNCH

LANDING

TRAFFIC CONTROL

TURNAROUND

TERMINAL

ASSEMBLY/INTEGRATION

EXPENDABLEELEMENTS

22

KABOM BasicsResources

• The spaceport has R resources– Each resource r has a set r of assigned activities.

– Each resource r has a per unit of capacity fixed operational cost, fcr and one time acquisition cost, acr.

– The capacity for a resource r, xr, is estimated by:

xr. = Round up(F x for all a r [E’(pa)] / T)

F = Round up(D/VC)

where

T = Time of spaceport operation F = Number of flights per year D = Annual demand (lbs.) VC = Vehicle per flight capacity

23

KABOM BasicsFleet Size

• The system requires a fleet of Vn vehicles to satisfy annual demand– The process network and average expected

activity times are used to determine the set of activities in the critical path .

Vct = for all a [E’(pa)]

Vfr = T / (Vct + Vot )

Vn = Roundup (F / Vfr)

where Vct = Vehicle ground cycle time Vot = Vehicle

orbit time

24

KABOM BasicsArchitecture

User InterfaceActivity Generator

Report Generator

Activity Library

User

Cost Generator

Improvement Agent

Process Model/Simulation

Critical Path/ Resource Capacity Generator (Process Model)

25

KABOM BasicsSimulation Model

• Why the use of simulation:– Better representation of a dynamic

environment.– Estimate bottlenecks, resource utilization,

queues, ..., costs (VS Team objective 1).– Provide designers a graphical (animation)

representation of their system at work.. (Team objective 2)

– Allows the evaluation of scenarios. • For example, a version of the model could be

developed where two types of vehicles can be modeled (I.e. Shuttle and Venture Star).

26

KABOM BasicsImprovement Agent and Cost

Calculations• Improvement Agent

– Heuristic to improve total performance. • The heuristics will recommend changes to resource

capacity and/or fleet size parameters. Several iterations are conducted (simulation repeated) until cost improvements are not obtained.

• Cost calculations– Activity costs tracked during the simulation

run.– Total costs include activity costs and resource

related costs– Vehicle acquisition costs are included as an

input from the designers.

27

Outline for this section

• Introduction• Previous work

• KABOM• Project Extensions

• Project Status

28

Project Extensions

• Approach used to develop design analysis tools for manufactured products (an advanced ABC analysis)– Help product designers estimate costs/processes– Evaluate what if scenarios in a design (mfg

processes)– A “real time” version of the model could

estimate the effect of a new product on current system load.

• Order for “customized product” arrives• Tool determines activities, costs, ..• Order is added to current production schedule • Decision on taking the order or how to schedule it

29

Outline for this section

• Introduction• Previous work

• KABOM• Project Extensions• Project Status

30

Project Status

• A prototype was developed last summer while at KSC

• Student project on a web based version of the AATe

• Grant proposal - excellent evaluations, waiting for funding.

31

PRESENTATION OUTLINE

• Describe an “interesting” research project

• Present an overview of current research and

teaching interests/projects

• Discuss my vision for the future

32

Overview Of Research/ Teaching Interests And

Projects

• Research– Production Planning– Supply Chain Management– IS/OM/Simulation Interaction– Space operations

• Teaching– Games– Excel applications – Advanced POM courses/ APICS/ ERP

(Future)

33

Production Planning Research

• Cellular manufacturing systems– Development of tools and models to plan/schedule

on systems where production resources are organized as cells and products into common families.

– Work with Jeet Gupta from Ball State University– Paper in progress

• Parallel Machine Scheduling with Multiple Criteria– Development of heuristics for difficult multi-criteria

problems in the parallel machine setting.– Work with Jeet Gupta and Kazou Nakatani (FGCU)– Two papers accepted for publication, several in

progress.

34

Supply Chain Management Research

• Supplier Decision Making– Evaluation of the role transportation plays on

supplier decision making.– Work with John Tyworth from Penn State

University– One paper accepted for publication

• Supply Chain Simulation– Development of simulation models for supply

chain environments in order to evaluate transportation and scheduling priorities..

– Work with John Tyworth

35

Interaction of IS, OM, and simulation

• Supply Chain Management– Development of architecture to support “real time” DM..– Utilizes simulation to investigate “what if” scenarios.– Work with J. Tyworth and K. Nakatani.

• Project Management– Development of architecture to support Web based

project management. – Utilizes simulation to evaluate the CP and effect of “what

if” scenarios.– Prototype developed and tested at NASA.– Work with K. Nakatani.

36

OM Games/Exercises

• Development/application of class games to enhance student learning– Beer game from Jacobs web page– In-class exercises after “lecture”.

• QFD - development of a QFD diagram for a business class

• SPC - process simulation

– JIT game with Lego/remote controlled cars (Rokenbok)

37

Excel Applications

• Use of Excel for operations analysis– Data analysis/ “real time” behavior graphs– Queue models– Forecasting– Capacity Analysis

• Use of Excel for quantitative methods course– Math Programming– VB front end– “Real world” application development

38

PRESENTATION OUTLINE

• Describe an “interesting” research project

• Present an overview of current research and teaching

interests/projects

• Discuss my vision for the future

39

My vision of the futureresearch

• Continue research on all four areas– SS: applications of KABOM to manufacturing,

IS/Simulation, Production Planning, SCM– Publication in leading journals, – Long term - editor of leading journal

• Develop relationships with manufacturers in El Paso and Juarez.

• Manufacturing-Logistics Center? - outreach program.– Foster knowledge creation in the ML field - faculty and

student projects, grants, ....– Foster business2business partnerships/creation of new

businesses– Continuing education programs?

40

My vision of the futureteaching

• Teaching/POM– Continue development of teaching games

• Program/curricula development– Graduate program in Manufacturing

Logistics?– Integration of APICS certification into

some undergraduate and graduate courses?

– Revision of POM undergraduate program?– ERP integration to the POM program?

• Link undergraduate POM students to faculty research projects

41

•Thanks for your time

•Questions?