R O L E   O F   T H E   C H I E F   S C I E N T I S T

WE WANT TO AVOID THESE TYPES OF PROBLEMS:

Course Outline Challenger Disaster Disclaimer National Security Space Programs Mythical Man – Month Resolving Engineer – Mgr Conflicts Analytic & Gaming Sims Complex Systems

NATIONAL SECURITY SPACE   •   Primary driver cost, not mission success   •   Unrealistic budgets, unexecutable programs   •   Competitor not “burdened” by actual cost   •   Budgets match unrealistic estimates   •   Requirements creep   •   Cost growth 50% – 100%

SHUTTLE

•   Success – oriented schedule •   Top – down testing         – Ok for design         – Not for test •   Redefining success         – 1/3 burn-through         – Order of magnitude estimates

•   Hiding failures •   Ignoring difficult issues •   Punishing those who notice problems •   Marketing over engineering

O N E   S O L U T I O N – DEFINE A ROLE —

•   Person with clout •   Ensure these suggestions are followed •   Loyalty is to technical credibility     — Not to marketing •   “The buck stops here”     — Catch mistakes     — Catch deliberately misleading work     — Ensure no “1 in 100,000” whoppers •   Duties vary with program size     — To avoid excessive overhead     — Most tasks should be done anyway         •   Office ensures not ignored         •   Need experienced person         •   Need sufficiently powerful office

•   Reports/Brfs to Outside Customers       — Chief Scientist examines these       — Works w developers from start       — Ensures meet tech stds           •   1:   No "whoppers"           •   2:   Follow good SE practices           •   3:   Address customer needs           •   4:   Follow std prog processes   •   Products to Outside Customers       — Chief Scientist examines these       — Works w developers from start       — Ensures meet tech stds   •   Does NOT report to Program Mgr

TOP

POWER ~ P R O G R A M   M A N A G E R

ENSURES MEETS TECHNICAL STANDARDS

1:   No "whoppers" 2:   Follow good SE practices

3:   Address customer needs

<---   WHY IS THIS #3, NOT #1?

4:   Follow std program processes

•   Part of 2nd priority: customer needs   •   If you have whoppers       — You lose credibility with customer       — You lose credibility with the public   •   Your first priority is YOUR REPUTATION       — Think Space Shuttle

TOP

P R O B L E M   P R O J E C T S

Projects that have had       — Schedule slips       — Cost overruns       — Trouble meeting tech stds Programs likely to have problems       — Look across program       — Are there trends?       — Similar problems?       — Similar fixes?

•   Problem projects can receive attention       — Project manager can make request       — Program Manager can make request       — Chief Scientist can make request       — Critical mass of project engineers can           •   Critical mass defined at program start           •   Critical mass defined for each project           •   Project leaders can't hide problems

TOP

How many programs set up, in advance, procedures for dealing with problems? How many programs prefer to hope they’ll never have any?

TOP

C H I E F   S C I E N T I S T

RESOURCES

REQUIREMENTS

•   Need to have Office of Chief Scientist       — One person too few on big programs       — About 3% of program       — 30-person office, one Chief Scientist           •   Would work multiple projects           •   3% of each project's budget       — Smaller programs           •   Project leader           •   Work especially tough problems           •   Other projects as budget allows

•   Ph.D. in appropriate field       — Aero engineering       — Physics       — Math   •   20+ years experience       — Multiple programs       — Project leadership   •   ATF/TF/STF (Technical Ladder)   •   Management experience       — Project management experience       — Regular management experience           •   Not grounds for rejection           •   No substitute for tech expertise

A “tax” of 3% is small enough that it would not wreck contractor rates.   But it’s big enough to allow the Chief Scientist to actually do something.

TOP