||Ex-SE: Expert System on Systems Engineering
Alliant Techsystems' MK 50 Torpedo Program
(Risk Assessment /
The assessment team for the MK 50 torpedo program developed its technical
risk baseline early in the technical requirements phase. They used charts to
track actual risks against predicted risks. Alliant Techsystems used a
three-color system: red (R) for high risks, yellow (Y) for moderate risks, and
green (G) for low risks. The team assigned low risks when their practices
complied with the best practices in the DoD templates. Key best practices that
Alliant Techsystems used included design reviews, thermal analysis, and test,
analyze, and fix.
The associated figure shows MK 50 torpedo program's overall summary
indicators to track technical risks in design, test, production, and technical
management. Each torpedo section developed several risk indicators for design,
test, production, and technical management. These indicators were then summed
into overall indicators of risk for the section. Sections included sonar,
warhead and fleet exercise, command and control, propulsion, launch
accessories, tactical software, and automatic test equipment.
The risk indicators were summed into overall indicators of risk for the
section. For example, the figure shows the indicators the sonar section used.
The team analyzed these indicators as red (high risk), yellow (medium risk),
or green (low risk) and then totaled them to obtain a summary risk rating for
design, test, production, and technical management. The team conservatively
gave a yellow (medium risk) rating for design overall for the sonar section.
Although four of the design indicators were green (low risk), the
engineering-changes indicator was red (high risk) and the failure-activity
indicator was yellow (medium risk).
The engineering-changes indicator was red (high risk) because of the number
of unresolved changes that exceeded 120 days in age. Alliant Techsystems
tracked both the number of engineering changes per month and age of the
changes still to be resolved. They used red for unresolved changes averaging
more than 120 days in age, yellow (medium risk) for unresolved changes
averaging 30 to 119 days in age, and green (low risk) for unresolved changes
averaging less than 30 days in age.
The figure shows the number of engineering changes per month, the average
age in days, the number of changes unresolved after 30 days, and the number
unresolved after 120 days. In September and October 1988, the number of new
engineering changes was well below plan, but the backlog of unresolved changes
was still high. By October, the backlog consisted entirely of "old dogs"
unresolved changes with an average age of 130 days.
The figure shows a risk report in which the high and moderate alert curves
are superimposed over the actual total number of technical risks. These risks
were plotted monthly from January 1987 to February 1988.
Management focused not only on individual high-risk items and the number of
changes per month, but also on the cumulative number of open risk items over
time. Having too many technical risks was a technical risk itself because of
the difficulties in managing such a large number of them. The team
predetermined critical totals and drew alert curves for the cumulative risks.
Alert curve 1 signified a moderate technical risk. Alert curve 2 signified a
high technical risk. The team notified upper management if the total number of
risks exceeded these predetermined alert curves. Exceeding Alert 1 or Alert 2
would determine which level of management should review the technical risks
and assign corrective-action plans.