曝光台 注意防骗
网曝天猫店富美金盛家居专营店坑蒙拐骗欺诈消费者
Detailed Resource Allocation
Existing planning tools often underestimate resource requirements because;
they cannot model iterations and feedback loops and therefore do not account for the
necessary repetition of key activities
they operate at a high level of abstraction which can miss important detail
These weaknesses are also amplified by the tendency, observed by cognitive psychologists, for
humans to be overconfident in their judgements about their own capabilities, which has a pronounced,
yet largely unresearched impact on projects. 26 With respect to this overconfidence phenomenon, in
the projects we observed, defensive scheduling was often frowned upon and seen as somewhat
defeatist. These human factors issues may never be fully overcome but to help to counter this a
Planner is required that has a very sophisticated resource allocation model which automates resource
deployment to meet project objectives. The Planner must operate at a fine level of resolution, giving
users a realistic picture of the envisages tasks and therefore allowing a very detailed estimate of
resources required.
Sophisticated Risk Analysis
There is no doubt that poor risk management is undermining many contemporary engineering projects.
In aerospace projects many of the real risks seem to be adequately identified, but their severity is
often underestimated. Again, there is an overconfidence over how easily the risk can be mitigated,
particularly the time that will be needed to put the project back on track. But over and above this
human factors issue, the current tools do not help. It is both difficult and very labour intensive to
undertake sensitivity analysis and risk analysis using existing planning tools. This is because existing
tools do not have built-in statistical sampling or experimentation capabilities. Thus the planner is
required to manually perform “what-if” investigations or use safety management techniques like
failure mode and effects analysis (FMEA) that were actually developed to analyse components. The
PERT (Project Evaluation and Review Technique) multi-time estimate method often used by existing
tools is acknowledged to give optimistic estimates of project duration and uses statistical assumptions
that are now widely discredited. In particular, using PERT as part of the critical path method has the
decisive disadvantage that the timeline of tasks is given in a linear fashion thus masking the true
nature of cyclical activities.27
To overcome these problems the authors advocate a Planner employing a full Discrete-Event
Simulation (DES) within its kernel, making it very easy to undertake systematic and statistically
meaningful “what-if” analyses, which ensure that robust plans are generated. This approach allows
managers to see the impact of contingency plans and backup resources on project risk.
Accelerated Change Management
Irrespective of the quality of planning undertaken it is almost inevitable that projects will deviate from
the original plan due to a change in objectives or customer requirements. In both defence and
aerospace this is a common problem as customers frequently change their requirements. On civil
aircraft programmes launch customers have unprecedented power, as their order is vital to the whole
26 J.B. T. Evans, Bias in Human Reasoning Causes and Consequences, (Lawrence Erlbaum Associates, New
Jersey, 1989).
27 D. Grover, ‘Graphical Project Planninmg Techniques: an Overview of Gantt, PERT and CPM’, (2002)
@/www.egr.msu.edu/classes/ece480/goodman/ganttv1.pdf
Why Projects Fail
15
credibility of the project. In reality it is very difficult to deny their request for requirements changes.
But existing tools have great difficulty in reflecting these changes, particularly if the project manager
has failed to capture the underlying dependency logic. Generating new Gantt charts and re-allocating
responsibilities and resources are very time-consuming using conventional planning methods. But
change can be easy to handle if the following features are available in the Planner.
The use of a fully connected dependency network allowing new or changed deliverables to be
incorporated quickly and easily;
A wide and distributed network permitting re-planning to be widely devolved to domain
experts who are better able to analyse and plan the consequences of changes;
A high degree of automation which requires no manual intervention in rescheduling/resequencing
activities and allocation of resources;
Cost implications of changes to project objectives/ requirements/ deliverables able to be
calculated quickly and easily. These can be identified in an auditable manner to facilitate
中国航空网 www.aero.cn
航空翻译 www.aviation.cn
本文链接地址：
航空资料41(81)
