This is a very simplified view of what can happen with a safety- or otherwise mission-critical system
analysis. It isn??™t cheap. It requires a lot of thought on the part of engineers and it requires that every bit
be tested to prove that nothing bad will happen in the time frame of interest. It is a lot less expensive
than an accident, or losing several million dollars in a faulty trade.
TABLE 3.4
Cost of risk mitigation
Solutions to (causes of) risks Cost (US$)
Delay
(elapsed)
Hire a C# guru for six months 60,000 0
Send all development staff on a C# course 20,000 1
Rewrite entire system
a
250,000 8.5
Write and review proper requirements specifications including throughput 30,000 1
Limit voice input (optionally research a different voice input engine) 0 0
Plan to spend six tester-months for every developer-year spent on the project
(figure represents increase over original budget)
b
70,000 ??“4
Buy code quality tools 100,000 ??“1
Have critical features prototyped and performance-tested, so that any
potential shortcoming is known well-enough in advance
122,000 3
Total 652,000 8.5
a
If the system is so clunky that rewriting it is the better option, then the alleged elapsed delay of 8.5
months is far
less
than the time taken to demonstrate that the old system is unfixable. However this plan
is being made well before that unhappy point will be reached and the delay against the original (highlyunrealistic)
plan becomes evident.
Pages:
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153