POWER Enhancements for Modeling Edge Organizations

1
POW-ER Enhancements for Modeling Edge
Organizations

• Raymond Levitt and Marc Ramsey
• Edge Project Workshop 1
• Stanford University
• 3 December 2004

2
Motivation

• Model behavior of Edge organizations using
  enhanced project simulator
• Provide support for testing Edge organization
  hypotheses
• Improved modeling of micro-behaviors
• The flow of knowledge in Edge organizations, by
  adapting MS Inventory Theory
• The exploration of new knowledge and exploitation
  of old, based on sociological and social
  psychological theory (sense-making, trust, and
  collaborative networks).

3
Virtual Design Team (VDT)

• 15 years of ongoing research at Stanford
• Project-oriented organization simulator
• Models actors and tasks within project
• Emphasis on hidden work resulting from
  coordination and exception handling
• Unique insights for fast track projects with
  multiple parallel strings of tasks
• Commercialized as SimVision project design and
  management software

4
VDT Research History
More Innovative
More Innovative
More Innovative
More Innovative
More Innovative
More Innovative
Work Processes
Work Processes
Work Processes
Work Processes
Work Processes
Work Processes
VDT
-
1,2 Coordination Costs
VDT
-
1,2 Coordination Costs
90
-
94 Cohen/
90
-
94 Cohen/
Christiansen
Christiansen
VDT
-
3 Different Goal Emphases
VDT
-
3 Different Goal Emphases
95
-
99 Thomsen
95
-
99 Thomsen
Salazar
-
Kish
Salazar
-
Kish
More Impact of IT/
More Impact of IT/
More Impact of IT/
More Impact of IT/
More Impact of
More Impact of
More Impact of
More Impact of
More Impact of IT/
More Impact of
More Impact of IT/
More Impact of
Past
Past
Communication
Communication
Communication
Communication
Social Processes
Social Processes
Social Processes
Social Processes
Communication
Social Processes
Communication
Social Processes
technologies
technologies
technologies
technologies
Work
Work
VDT
-
4 Dynamic Work Processes
technologies
VDT
-
4 Dynamic Work Processes
technologies
96
-
99 Fridsma/
96
-
99 Fridsma/
Cheng
Cheng
Single Project
Single Project
Single Project
Single Project
VDT
-
5 Knowledge Networks
VDT
-
5 Knowledge Networks
99
-
02 Lambert/
99
-
02 Lambert/
Program
Program
Program
Program
Buettner
Buettner
Enterprise
Enterprise
Enterprise
Enterprise
Community
Community
Community
Community
of Practice
of Practice
of Practice
of Practice
6 Institutional Costs
VDT
-
6 Institutional Costs
VDT
-
Dynamic
Dynamic
Dynamic
Dynamic
Future
Ongoing
03
-
05 Buettner/Murray/
03
-
06 Buettner/ Mahalingam/
Hierarchies and
Hierarchies and
Hierarchies and
Hierarchies and
Mahalingam
Orr/Taylor/Horii/ Post Processors
Murray/KHosraviani
Networks
Networks
Networks
Networks
Work
Work
More Flexible
More Flexible
More Flexible
More Flexible
More Flexible
More Flexible
Organizations
Organizations
Organizations
Organizations
Organizations
Organizations
5
Example of VDT Model
6
VDT Task Model

• Specified amount of direct work volume
• Require a specific skill or knowledge
• Predetermined task complexity measures
• Uncertainty
• Requirement and solution complexity
• Relationships between tasks
• Successor dependencies
• Communication dependencies
• Rework dependencies
• Task has a single pre-assigned actor

7
VDT Actor Model

• Probabilistic info-processing behavioral model
• Defined set of skills/knowledge
• Skill levels are fixed for duration of simulation
• Actor represents an individual or team
• Fixed work capacity in FTEs
• Simulated work speed determined by capacity and
  actor rating for task-required skill
• Relationships between actors
• Supervisor
• Meetings
• Info-exchange

8
VDT Simulation (I)

• Actors receive requests in their inbox
• Direct task work and rework
• Task and project related coordination
• Exceptions and decisions
• Actors select requests based on heuristics
• Priority, FIFO, LIFO, and random
• Actors may become backlogged
• Slow responding to requests
• Requests may expire before being processed

9
VDT Simulation (II)

• Task work/rework may result in exceptions
• Functional exceptions
• Project exceptions
• Exception decision may result in additional
  rework
• A task is complete when all outstanding direct
  work and rework have been processed by actor
• Simulation complete when all tasks done
• Data from multiple simulation runs averaged to
  produce simulation results

10
Examples of Simulation Output
11
POW-ER Enhancements to VDT

• Knowledge flow
• Exception handling via knowledge networks
• Differing cultural/organizational background
• Trust relationships between actors
• Quality of communications

12
Knowledge Flow and K-Networks

• Based on work of Rahinah Ibrahim, Stanford
• Peer to peer knowledge relationships
• Models distributed problem solving
• Actors perception of other actors knowledge may
  differ from reality
• Learning from knowledge relationships
• Gain knowledge over time from interaction with
  others of higher skill/knowledge

13
Cultural/Organizational Background

• Based on work of Tamaki Horii, Stanford
• Model interacting actors from different cultural
  and/or organizational backgrounds
• Each actor has specified culture and/or
  organization
• Simulated behavior of actors may be specialized
  according to background
• Model cross cultural communication difficulties

14
Trust Relationships

• Based on work of Roxanne Zolin, NPS
• Models effects of perceived trustworthiness
  between actors on coordination
• Examine effect of trust on knowledge flow

15
Quality of Communications

• Models quality of communication based on
• Cultural and organizational differences
• Communication bandwidth
• Proximity of actors
• Presence of noisy or incorrect information

16
POW-ER Simulation Framework

• Model Editor
• Implemented as custom extension to Visio
• Simulation Engine
• Object-based implementation using Python dynamic
  programming language
• Data Handling
• XML format used for models and result data
• Chart and report generation
• Data can be exported to database or Excel

17
Open Issues Research Questions

• Modeling projects vs. ongoing processes
• How do we model shared awareness?
• Blackboard metaphor
• Peer to peer relationships
• Information overload issues
• What metrics can be used to evaluate performance
  of Edge organizations?