Lawrence Livermore National Laboratory Managing Complex Interdependencies Cynthia A. Wagner Manager, Office of Performance Excellence November 28, 2007.

Lawrence Livermore National Laboratory

Managing Complex Interdependencies

Cynthia A. Wagner

Manager, Office of Performance Excellence

November 28, 2007

The Emerging Challenge of Human Performance Improvement:

2Option:UCRL# Option:Additional Information


The Challenge

Barriers and Practices are Dynamic



Balancing Acts at Play

Dynamic and Multi-dimensional • Influence• Proficiency• Flexibility



Proactive Management

Human-System Interfaces• Knowledge of Current

Practices Barriers Outcomes

• Awareness of Emerging Changes

USER NEEDS: WHAT DOES WORK, DOESN’T WORK, and IS LIKELY TO CHANGE?



Causal Analysis

Structured, questioning process Enables recognition of practices and beliefs in an

organization, or does it? Why don’t we do more Root Causes?

USER NEEDS: DISCUSSION OF VALUES AND BELIEFS



Human Nature of Analysis

Does our preference for causal methods simply reflect our relationship with the tools?

Do our linear approaches oversimplify the complexity we face?

What might keep us from being willing to explore further?

USER NEEDS: ROBUST, SYSTEMATIC and SYSTEMIC



Convergence of Information and Thinking

Do Causal Teams Really Achieve Common Understanding?

How Can We Create Transparency and Traceability of the Sensing and Thinking Process?

USER NEEDS: TRANSPARENT and TRACEABLE



Human Limitations

We always know more than we can tell

We always tell more than we can write down



Making Sense for Others

Are expectations reasonable given the complexity of interdependencies?

Are traditional analysis reports effective for making sense of findings and creating buy-in?

USER NEEDS: EFFICIENT and EFFECTIVE COMMUNICATIONS



Summary of User Needs

Method• Robust, Systematic and Systemic• Easy to Learn and Use

Output• Knowledge and Insight• Traceable and Transparent Discussion• Effective and Efficient Communications



Consider Stream Analysis

Compatible with HPI• Open Systems Theory• Social Cognitive Theory

In Practice• INPO experience

Applicable across the organization at all levels• Project leaders, unit managers, organizational

advisors, oversight teams, business executives

“Stream Analysis - A Powerful Way to Diagnose and Manage Organizational Change” by Professor Jerry Porras, Stanford Graduate School of Business



Why Stream Analysis?

*M. Beer and N. Nohria, “Cracking the code of change” Harvard Business Review for turnaround, p 1 1997.

Human-System interfaces are not linear in nature

Complex interconnections between organizational components control and influence behavior, processes and performance

Enterprises spend billions on Improvement initiatives

70% of change initiatives by the fortune 100 fail *

Processes

Performance

Behavior



Interdependent Components

Core Structure (organization)

Social Factors (behaviors and values)

Technology (integrated work processes)

Physical Setting (environment)



Streams



Process Management

Divergence

Group Consensus

Review Process

Retrieval Method

Vertical Slices

Shared Assumptions

Meaning of Issues

Organization

Feedback Loops

Timing

Actions

Convergence



Stream Analysis provides a step by step procedure for:

Procedure

1. Forming Change Management Team2. Collecting Data3. Categorizing Problems4. Identifying Interconnections5. Analyzing the Problem Chart6. Formulating a Plan of Action7. Tracking the Intervention Process



Example of Discussion File

5.2.3 Inadequate Labeling

Labeling requirements provide a barrier that communicates the presence of potentially hazardous materials.

This barrier failed because workers and management did not implement ES&H labeling requirements.

Whether the NR-1 check source was a Class I sealed source or not, the level of radioactive material it contained qualified it for labeling requirements as specified in ES&H Manual Document 20.2.

Had the NR-1 check source been accurately labeled, it would have been clearer that additional controls applied, such as being in an inventory, periodic swiping, and storage.

In addition, labeling-related deficiencies from the 2003 Radiation Protection Assessment were closed without being fully corrected



Comparison to User Needs

Systematic and Systemic• 4 Streams represent the system• Software aids execution of the process

Transparent and Traceable• Discussions remained fact-based. Assumptions and

questions were captured for reference. • Binning provides a self-check on understanding of

the issues. • Diagnostics captures the logic used

Effective and Efficient Communications• Creates a “Rich” Picture



Case Study

Incident Analysis on Contamination • 13 Judgments of Need

Investigation summarized:• Several of the conclusions of the IA Committee involve the failure of

controls associated with sealed sources. • This is because the NR-1 check source was assumed to have been a

sealed source at the time the contamination began to spread. • Had the controls for sealed sources been applied, the IA Committee

believes the contamination would have been detected before being spread to other facilities and off-site. However, the reader is urged to remember that the real core of this incident was the handling of a legacy item.



Case Study

Stream Analysis Results• Unclear Roles and Responsibilities

People “jumped the turnstiles”• Process was Error Prone

Error Traps for Unidentified Sources• Flawed assumptions

Lack of Questioning Attitude Safety was not first



Stream 1



Diagnostic



Theme – Ineffective Characterization



Theme – Reliance on Others



Stream 2



Stream 2



Theme on Safety Culture



Theme – Safety Culture 2



Questions?

Lawrence Livermore National Laboratory Managing Complex Interdependencies Cynthia A. Wagner Manager, Office of Performance Excellence November 28, 2007.

Documents

dynamic slide

traceable slide

systemic slide

effective communications

efficient communications

traditional analysis

stream analysis compatible

causal teams